(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Proceedings of the American Society of Agronomy"

OAK ST. HDSP 



THE UNIVERSITY 
OF ILLINOIS 
LIBRARY 

v.9cop.o 

mmm^ 



Digitized by the Internet Archive 
in 2013 



http://archive.org/details/proceedingsofanne9191anner 



JOURNAI. 

OF TIIIC 

AMERICAN SOCIETY 
OE AGRONOMY 



VOLUME 9 



1917 



PUBLISHED BY THE SOCIETY 



PRESS OF 
THE NEW ERA PRINTING COMPANY 
LANCASTER. PA. 



DATES OF ISSUE. 

Pages 1-48, January 15, 191 7. 
Pages 49-96, February 15, 191 7. 
Pages 97-144, March 28, 191 7. 
Pages 145-200, April 25, 1917. 
Pages 201-256, May 21, 191 7. 
Pages 257-304, September 25, 191 7. 
Pages 305-352, October 22, 191 7. 
Pages 353-384, November 20, 1917. 
Pages 385-432, December 27, 1917. 



ERRATA. 



On page 48, second paragraph, lines 6 and 7, read "$398,000, of which 
$250,000 is for the eradication of citrus canker and the ..." 

On pages 290 and 291, in Tables i and 2, read " (NH4)2SO.i throughout in 
place of " (NH40H).S04." 

On page 326, Table i, first line, read "June 10" for "June i." 



iv 



CONTENTS 



. No. I. JANUARY. 

Pace. 

Le Clerc, J. A., and Bailky, L. H. — The Composition of Grain Sorghum 

Kernels i 

HuTCHESON, T. B., and Quantz, K. E. — The Effect of Greenhouse Tem- 
peratures on the Growth of Small Grains (PI. i and 2 and Fig. i). . 17 
Salmon, S. C. — The Relation of Winter Temperatures to the Distribution 

of -Winter and Spring Grain in the United States (Figs. 2 and 3)--- 21 
Skinner, J. J., and Beattie, J. H. — Influence of Fertilizers and Soil 

Amendments on Soil Acidity 25 

W^aller, a. E. — A Method for Determining the Percentage of Self-Pollina- 

tion in Maize 35 

Shaw, Charles F. — A Method for Determining the Volume Weight of 

Soils in Field Condition 38 

Agronomic Affairs. 

The Society in 1917 43 

Annual Dues of Members 44 

Membership Changes 45 

Notes and News 46 

No. 2. FEBRUARY. 

Call, L. E., and Sewell, M. C. — The Soil Mulch 49 

Pieters, a. J. — Green Manuring: A Review of the American Experiment 

Station Literature — i 62 

Hopkins, Cyril G. — A Limestone Tester (Fig. 4) 82 

Agronomic Affairs. 

New Books 90 

Membership Changes 91 

Notes and News 92 

Directory of Local Sections 96 

No. 3. MARCH. 

FiPPiN, Elmer O. — Livestock and the Maintenance of Organic Matter in 

the Soil (Fig. 5) 97 

Montgomery, E. G. — Heating Seed Rooms to Destroy Insects (Fig. 6) 105 

PiETERS, A. J, — Green Manuring: A Review of the American Experiment 

Station Literature — 2 109 

Arny, A. C, and Thatcher, R. W.— The Effect of Different Methods of 
Inoculation on the Yield and Protein Content of Alfalfa and Sweet 
Clover — 2 127 

McCall, a. G. — A New Method for Harvesting Small Grain and Grass 

Plots (Figs. 7 and 8) 138 

V 



vi 



CONTENTS. 



Page. 



Agronomic Affairs. 

Membership Dues 141 

Membership Changes 141 

Notes and News 143 

No. 4. APRIL. 

Davidson, J., and Le Clerc, J. A. — The Effect of Sodium Nitrate Applied 
at Different Stages of Growth on the Yield, Composition and 

Quality of Wheat 145 

Bailey, C. H. — The Quality of Western-Grown Spring Wheat 155 

PiETERS, A. J.— Green Manuring : A Review of the American Experiment 

Station Literature — 3 162 

Waller, Adolph, and Thatcher, L. E. — Improved Technique in Prevent- 
ing Access of Stray Pollen (PI. 3) ' 191 

Agronomic Affairs. 

New Books 196 

Membership Changes 199 

Notes and News 200 

No.. 5. MAY. 

Grantham, A. E. — The Relation of Cob to Other Ear Characters in 

Corn (PI. 4) 201 

A-iCHER, L. C. — Whole vs. Cut Potato Tubers for Planting on Irrigated 

Land— I (Pis. 5 and 6) 217 

Welch, John S. — Whole vs. Cut Potato Tubers for Planting on Irrigated 

Land — 2 • 224 

Boshnakian, S. — The Comparative Efficiency of Indexes of Density, and 
a New Coefficient for Measuring Squareheadedness in Wheat (PI. 

7 and Figs. 11-15) 231 

Bailey, C. H. — The Moisture Content of Heating Wheat 248 

Agronomic Affairs. 

New Books 252 

No Summer Issues of the Journal 253 

Membership Changes 254 

Notes and News 255 

No. 6. SEPTEMBER. 

Jardine, W. M. — A New Wheat for Kansas 257 

Halsted, B. D., and Owen, Earle J. — Influence of Position of Grain on the 

Cob on the Growth of Maize Seedlings 267 

Bailey, C. H. — The Handling and Storage of Spring Wheat (Figs. 17-20). 275 
A Committee of the Minnesota Section. — The Color Classification of 

Wheat 28T 

Kelley, W. p. — The Action of Precipitated Magnesium Carbonate on Soils 285 

Conner, S. D. — Excess Soluble Salts in Humid Soils 297 

Agronomic Affairs. 

The Annual Meeting 302 

Membership Changes 302 

Notes and News 303 



("ONTKNTS. \ii 

I'Af.Ii. 

No. 7. OCTOIUCR. 

SwANSON, C. O.— The VA'U\-[ of I'roloiiKc'd (Irovvini; of Alfalfa 011 tlu- 
Nitrogen Content of the Soil 

LovK, H. H.. and Wentz, J. B. — Correlations between Kar C'liaracters and 

Yield in Corn 3i5 

DuNNEWALD, T. J. — Vegetation on Swamps and Marshes as an Indicator 

of the Quality of Peat Soil for Cultivation 322 

Gericke, W. F. — Some Effects of Successive Cropping to Barley 325 

Schuster, Geo. L. — A Study of Soil Solutions by Means of a Semi-perme- 
able Membrane Supported by a Porous Clay Plate (PI. 8 and 
Figs. 21-23) 333 

Grantham, A. E. — The Relation of the Vigor of the Corn Plant to 

Yield 340 

Thatcher, R. W., and Arny, A. C— The Effect of Different Rotation 

Systems and of Fertilizers on the Protein Content of Oats 344 

Agronomic Affairs. 

The Tenth Annual Meeting 349 

Membership Changes 349 

Notes and News 350 

No. 8. NOVEMBER. 

Salmon, S. C. — Why Cereals Winterkill 353 

CoE, H. S. — An Annual Variety of Melilotus alba 380 

Agronomic Affairs. 

Membership Changes 383 

Notes and News 383 

No. 9. DECEMBER. 

Jardine, W. M. — The Agronomist of the Future (Presidential Address) . . . 385 

Agronomic Affairs. 

Report of the Secretary for 1917 391 

Funds Collected by the Secretary 391 

Meetings 393 

Local Sections 393 

Membership 393 

Journal and Proceedings 39S 

Minutes of the Annual Meeting . . .• 397 

Report of the Treasurer . 400 

Reports of Committees 401 

Executive 401 

Standardization of Field Experiments... 402 

Varietal Nomenclature 419 

Index 428 



JOURNAL 

OF THE 

American Society of Agronomy 



Vol. 9. January, 1917. No. i. 



THE COMPOSITION OF GRAIN SORGHUM KERNELS.^ 

J. A. Le Clerc and L. H. Bailey. 

INTRODUCTION. 

This paper gives the average results of a large number of analyses 
of the seed of the grain sorghums. The samples analyzed were from 
crops grown in the Panhandle of Texas by the Office of Cereal Inves- 
tigations of the Bureau of Plant Industry, U. S. Department of Agri- 
culture, in the five years from 1908 to 1912, inclusive. The investi- 
gation was undertaken for the purpose of ascertaining how the com- 
position of these grain sorghums varies in different years and whether 
such variation is a result of the climatic conditions prevailing during 
the growing and pregrowing periods. In addition to giving the aver- 
age composition of those grain sorghums that have been grown quite 
extensively, this paper contains analyses of a small number of sam- 
ples of the grain of certain' sorghums which are not so well known — 
shallu and broomcorn. There are also reported results of analyses 
of bread made in part from grain sorghum meal. 

The chemical analyses here given when compared with analyses of 
the ordinary cereals may serve as an index to the food value of the 
grain sorghums. The results show that these grains may be favor- 
ably likened to the cereals in food value. A general knowledge of 
this fact should serve to stimulate and to encourage a wider use of the 
sorghums for human consumption, as well as for stock feeding. 

1 Contribution from the Laboratory of Plant Chemistry, Bureau of Chem- 
istry, United States Department of Agriculture. Published by permission of 
the Secretary of Agriculture. Received for publication November i, 1916. 

I 



2 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

THE GRAIN SORGHUMS. 

The grain sorghums, which are indigenous to Africa and India, 
are of great antiquity and are grown extensively in those countries 
and China. They belong botanically to the general classification of 
sorghums (Andropogon sorghum) (3).^ Among them may be men- 
tioned milo, kafir, durra, and kaoliang, and crosses or hybrids among 
these. They are drought-resistant plants, and it is due to this fact 
more than to any other that they may be grown successfully and prof- 
itably in the dry-land region of the Great Plains. The cultivation in 
this region of such grains as milo and kafir is highly desirable, since 
the yield is greater than can be obtained from almost any other crop. 
The actual yield in dry-land areas is sometimes as high as 50 or even 
75 bushels per acre (5). 

The sorghum belt of this country is the southern part of the Great 
Plains area. It is nearly 400 miles wide and 1,000 miles long (4). 
The grain sorghums are grown to such an extent in the United States, 
— principally within the area peculiarly adapted to their growth, — 
that in 1910 over 3,000,000 acres were devoted to their culture, the 
crop having a value of about $30,000,000. 

Where grown in the Old World the grain sorghums are commonly 
used as human food and, indeed, often furnish the chief article of 
diet. In the United States they are generally employed as stock and 
poultry feeds, for which purpose they have been successfully used, as 
they are very similar in composition to corn. Recently attempts have 
been made in this country to subject sorghum grain to the process of 
milling and to employ the ground grain in the same way that corn- 
meal is used. Inasmuch as these grains do not contain gluten as such, 
meal made from them can not be used alone, as can wheat flour, for 
making yeast-risen bread. There seems to be no reason, however, 
why such meal should not be used as a partial substitute for wheat 
flour in the making of bread. Undoubtedly it can be used in making 
batter cakes and similar products. 

Milo, kafir, and the other grain sorghums might be used with profit 
for the manufacture of breakfast food and for popping, as substi- 
tutes for popcorn (6). Moreover, since these grains contain prac- 
tically the same proportion of carbohydrates as does maize, they can 
perhaps be profitably employed for the manufacture of alcohol or for 
the manufacture of a sirup similar to glucose or corn sirup. This 
suggestion is made as a result of the work done in Oklahoma by Baird 
(i) and by Baird and Francis (2). Baird states that kafir grain 

2 Reference is made by number to " Literature cited," p. 16. 



LRCLKRC IIATI.KV : COM POSITION OK CKAIN SOKCIHUMS. 3 

compares favorably with maize in digestibility, as shown by digestion 
experiments with chickens, in composition as shown by chemical 
analysis, and in suitability for the production of alcohol or glucose. 
His results also show that a crop of kafir removes less plant food 
from the soil than does a crop of maize. 

METHODS OF ANALYSIS. 

The methods of analysis used in this investigation were those given 
in Bureau of Chemistry Bulletin 107, revised, and known as the 
methods of the Association of Official Agricultural Chemists ( 7) The 
weight per bushel was obtained by the use of a miniature bushel or 
grain tester. The weight per 1,000 kernels was obtained by counting 
from each sample lots containing 200 and 300 kernels, respectively, 
and weighing them. The agreement should be within half a gram 
per 1,000 kernels. 

RESULTS OF BREAD-MAKING TEST. 

From experiments in the use of grain-sorghum meals as part 
substitutes for flour in bread-making, it appears that it is possible 
to make from a mixture of 25 percent milo, kafir, or other grain- 
sorghum meal, and 75 percent of good wheat flour, a very pleasing 
loaf of bread. Such bread has essentially the same composition as 
bread made from wheat flour, except that it contains somewhat more 
ash and fiber than does wheat flour bread. The striking difference 
between these sorghum breads and bread made from wheat flour alone 
is in the color, that from the sorghum-wheat mixture resembling 
somewhat a loaf made from rye-wheat mixture. Table i shows the 
composition of bread made with wheat flour alone and of bread in 
which milo, kafir, or kaoliang meal has been substituted for 25 percent 
of the flour.* 



Table i. — Analyses of bread made from wheat flour alone and from mixed 
flours containing 25 percent of grain-sorghum meal. 



Kind of bread. 


Water. 


Ash. 


Protein. 


Fat. 


Fiber. 




Percent 


Percent 


Percent 


Percent 


Percent 




35.00 


1.26 


9.60 


1.88 


0.13 


Milo and wheat 


32.76 


1.46 


9.25 




.35 


Kafir and wheat 


30.15 


1.66 


9.78 




.36 


KaoHang and wheat 


35-70 


1-43 


8.94 




.33 



3 The nitrogen determinations were made in the Nitrogen Laboratory, Bureau 
of Chemistry, under the supervision of T. C. Trescot. 

The baking experiments were carried on by Miss H. L. Wessling, of the 
Laboratory of Plant Chemistry, who also did the analytical work on the breads. 



4 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

RESULTS AND DISCUSSION OF ANALYSES. 

Of the analyses of grain sorghums only the minimum, maximum, 
and average results are given for each kind of grain for the series of 
years during which it was grown. These figures represent hundreds 
of analyses made during a period of five years, from 1908 to 1912, 
inclusive. It is believed, therefore, that they will give a very clear 
idea of the possible variation in the composition of sorghum grains 
grown in a particular section of the country and will show to what 
extent the composition has been affected by climatic conditions. 

Table 2 shows the minimum and maximum results and the average 
composition of the various grain sorghums for the series of years 
during which they were grown at Amarillo, Tex. The data are sum- 
marized in the last section of the table. 



Table 2. — Minimum, maximum, and average composition of various grain sor- 
ghums grown at Amarillo, Tex., 1908-19 12. 

DURRA. 



Variety and analysis. 


Water. 


Ash. 


Protein 
(iVX6.2S). 


Fat. 


Fiber. 


Carbo- 
hydrates. 


Weight 
per 1,000 
kernels. 


Weight 

per 
bushel. 




Percent 


Percent 


Percent 


Percent 


Percent 


Percent 


Grams 


Pounds 


White, 35 samples, 


















1908-1912. 




















8.05 


1.44 


11.25 


2.22 


1.06 


68.08 


20.7 


51.9 


Maximum 


10.46 


2.06 


15-19 


4.90 


3-39 


72.54 


32.0 


60.5 


Average 


9.49 


1.74 


13.69 


3-52 


1.48 


70.08 


28.2 


56.4 


Buff, 4 samples. 


















1908-1909. 




















8.56 


1.53 


12.06 


2.26 


1.41 


69.10 


20.1 


58.6 


Maximum 


10.58 


1.82 


15-00 


4.26 


1.80 


73.14 


34-5 


59-9 




9.60 


1.69 


13-06 


3-05 


1.62 


71.24 


28.1 


59.0 



DURRA KAFIR. 



Blackhull. 


















28 samples. 


















Minimum 


8.18 


1.76 


12.44 


2.92 


1-32 


65-42 


17.6 


46.7 


Maximum 


10.64 


2.19 


16.75 


4.46- 


2.25 


70.75 


24.5 


57.9 


Average 


9.49 


1.98 


14-38 


3.56 


1,70 


68.84 


22.3 


55-1 


Palehull, 


















19 samples. 


















Minimum 


8.62 


1-79 


11.69 


2.80 


1.24 


67.05 


17-3 


55.0 


Maximum 


10.56 


2.24 


16.06 


3-85 


2.36 


71.81 


27.9 


57.7 




9-79 


1.90 


14.88 


3-43 


1-57 


69.42 


22.3 


56.2 



LECLERC & bailey: COMPOSITION OF GRAIN SORGHUMS. 



5 



Table 2. — Miftimum, maximum, and average composition of grain sorghums 
grown at Amarillo, Tex., igo8-igi2. — Continued. 



KAFIR, 



Variety and analysis. 


Water. 


Ash. 


Protein 


Fat. 


Fiber. 


Carbo- 
hydrates. 


Weight 

per 1,000 
kernels. 


Weight 

per 
bushel. 




Percent 


Percent 


Percent 


Percent 


Percent 


Percent 


Grams 


Pounds 


Blackhull, 


















83 samples, 


















1908-T912. 


















Minimum 


8.50 


1-53 


11.25 


2.85 


1.30 


6594 


13-6 


55-5 


Maximum 


10.65 


1.99 


15-94 


4.00 


1.89 


72,79 


29.4 


60.3 




9.60 


1.78 


14.00 


3.44 


1.59 


69.52 


20.9 


59-0 


Red, 67 samples, 


















1908-1912. 




















8.63 


1.52 


9.69 


2.56 


1.22 


69.08 


14.7 


53-6 




10.62 


1.99 


14.40 


3.65 


1.82 


74.54 


29.7 


60.2 




9.71 


1.74 


12.37 


3-14 


T A Q 
I .40 


71.54 


20.3 


58.2 


White, 19 samples. 


















1908-1912. 


















Minimum 


8.76 


1. 55 


11.31 


2.78 


I. 18 


70.04 


15-7 


55-6 


Maximum 


10.88 


1.90 


13-50 


3.66 


1.79 


73-07 


30.1 


59-1 




9.84 


1.67 


12.44 


3-30 


1.50 


71.46 


23.1 


57-4 


Dwarf, II samples, 


















1908-1909 and 


















1911-1912. 


















Minimum 


9.38 


1.57 


II. 31 


3-01 


1-33 


69-59 


15-2 


58.3 




11.03 


1.72 


13.87 


3-55 


1.64 


72.65 


22.1 


59.6 




10.09 


1.66 


12.81 


3.19 


1.49 


70.81 


17-5 


59.2 



KAOLIANG. 



White, II samples, 


















1908-1912, 




















8.70 


1.58 


11.50 


4-07 


1,09 


68.85 


14-5 


55.8 




10.09 


1.99 


14.56 


5-21 


1-56 


71.87 


24.1 


59-4 




9.18 


1. 81 


13-56 


4-91 


1,26 


69-34 


19.7 


57.7 


Blackhull, 


















9 samples, 


















1908-1912. 


















Minimum. ..... 


8.79 


1.63 


II. 13 


3-39 


.96 


67-59 


15.4 


55-0 


Maximum 


10.24 


2.15 


15.63 


4-69 


1-59 


73.22 


23.5 


58.5 




9.48 


1.88 


13-06 


3-97 


1,26 


70.38 


18.6 


56.8 


Brown, 72 samples. 


















1908-1912. 


















Minimum 


8.25 


1.53 


10.25 


3-07 


1.06 


65.85 


13.7 


50.3 


Maximum 


11.07 


3-06 


15.31 


5.36 


2.58 


73.22 


30.3 


58.0 




9-38 


1.87 


13.42 


4,16 


1.39 


69.84 


19.2 


56.0 



MILO, 



Standard, 


















68 samples. 


















1908-1912. 


















Minimum 


8.02 


1-33 


9.81 


2,44 


1.26 


69-53 


27.8 


56.4 


Maximum 


11.27 


1.84 


14-13 


3-88 


1.74 


75-22 


39-4 


59-6 




9-33 


1.63 


12,63 


•3-16 


1-49 


71.86 


36.2 


58.1 



w 



6 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Table 2. — Minimum, maximum, and average composition of grain sorghums 
grown at Amarillo, Tex., igo8-igi2. — Concluded. 



MILO. 



V3.ricty 3.nci 3,n3.1ysis, 


Water 


Ash. 


Protein 
(A^X6.25). 


Fat, 


Fiber. 


Carbo- 
hydrates. 

1 


Weight 
per i^ooo 
kernels. 


Weight 

per 
bushel. 




Percent 


Percent 


Percent 


Percent 


Percent 


Percent 


Grams 


Pounds 


Dwarf, 55 samples, 


















1908-1912. 


















Minimum 


8. II 


1-43 


9.19 


2.85 


1.27 


69.23 


25.1 


55-0 


Maximum 


10.73 


1.83 


14-13 


3-78 


1.82 


75-33 


36.4 


59.6 


Average 


9.38 


1.63 


12.06 


3-27 


1.47 


72.20 


31.4 


58.2 


White, 19 samples. 


















1910-1912. 




















8.24 


I-5I 


9-63 


2.80 


1-39 


69-71 


25.2 


55-2 


Maximum 


10.45 


1.84 


14.44 


3-67 


1.86 


74-45 


35.5 


58.5 


Average 


9-35 


1.66 


12.88 


3-10 


1.53 


72.06 


32.9 


57.6 


Hybrid, 6 samples. 


















1911-1912. 


















Minimum 


9.21 


1-55 


12.13 


2.47 


1-59 


65.90 


27.1 


500 


Maximum 


10.82 


2.29 


15-50 


3-24 


3-57 


72.15 


36.2 


58.5 


Average 


10.03 


1.83 


14-13 


2.91 


2.21 


68.97 


32.4 


56.1 


SUMMARY. 


Durra, 39 samples. 




















8.05 


1.44 


11.25 


2.22 


1.06 


66.53 


20.1 


51.9 


Maximum 


10.58 


2.06 


15-19 


4-90 


3-39 


73-14 


37-3 


60. 5 


Average 


9-50 


1-73 


13-63 


3-47 


1-49 


70.30 


28.2 


56.7 


Durra kafir. 


















47 samples. 


















Minimum 


8.18 


1.76 


11.69 


2.80 


1.24 


65-42 


17.3 


54-1 


Maximum 


10.64 


2.24 


16.75 


4-46 


2.36 


71.81 


27.9 


57-7 




9.61 


1.95 


14.19 


3-51 


1.65 


69.08 


22.3 


55.6 


Kafir, 182 samples. 


















Minimum 


8.50 


1.52 


9-69 


2.56 


1. 18 


65-94 


13.6 


53-6 


Maximum 


11.03 


1.99 


15-94 


4.00 


1.89 


74-34 


30.1 


60.2 


Average 


9.70 


1. 75 


13-13 


3-30 


1-54 


70.30 


20.7 


58.2 


Kaoliang, 


















92 samples. 


















Minimum 


8.25 


1-53 


10.25 


3-07 


0.96 


65.86 


13.7 


50.3 


Maximum 


11.07 


3.06 


15-63 


5.36 


2.58 


73-22 


30.3 


59-4 


Average 


9-37 


1.87 


13-37 


4.23 


1.36 


69.84 


19.2 


56.3 


Milo, 150 samples. 




















8.02 


1-33 


9.19 


2.44 


1.26 


65.90 


25.1 


50.0 


Maximum 


11.27 


2.29 


15-50 


3-88 


3-57 


75-33 


39-4 


59-6 


Average 


9.39 


1.64 


12.50 


3-18 


1.52 


71.88 


33-9 


58.0 



The protein of grain sorghums is the most valuable constituent of 
these grains, although not the most abundant. So far as this sub- 
stance is concerned, sorghums have a composition approximating that 
of hard wheat. On the other hand sorghums have somewhat less 
fiber but more fat than wheat, while the proportion of ash is about 



LECLERC & 15AILEV : COM I'OSITION OF CRAIN SC)R(jIlUMS. 



7 



the same in both. Sorghum kernels usually average smaller than 
wheat kernels and weigh a little less per bushel. The main difference 
between the composition of sorghum and of wheat lies in the fact that 
the protein is quite unlike that of wheat. The ])rotcin of wheat is 
chiefly gluten, which possesses the characteristics desirable in bread 
making, while there is no gluten in the sorghum grain. 

An examination of the summary at the end of Table 2 shows that 
the weight per 1,000 kernels of the milos is considerably greater than 
that of the other sorghums, while the percentage of fat, ash, and 
protein is somewhat less. In composition the kafirs and durras lie 
between the milos and durra kafirs. The durra kafirs are the richest 
in protein, ash, and fiber, and have the lowest proportion of carbohy- 
drates and the lowest weight per bushel. Kaoliangs, on the other 
hand, contain appreciably more fat but less fiber than do the other 
grain sorghums and average the lowest in weight per 1,000 kernels. 

When the average compositions of the various grains for each of 
the five years from 1908 to 191 2 are compared the following ex- 
tremes are shown : 

Lowest average for Highest average for 

one year, one year. 

Water, percent 9.2 lo.o 

Ash, percent 1.6 2.0 

Fat, percent 3.0 4.9 

Fiber, percent 1.2 2.0 

Protein, percent 12.1 14.9 

Carbohydrates, percent 69.0 72.0 

Weight per 1,000 kernels, grams 17.5 36.2 

Weight per bushel, pounds 55.1 59.2 

The figures just given are averages for certain years in the series 
of years. In the case of individual results the actual variation from 
the minimum to the maximum is much greater, as may be seen from 
the following figures taken from the summary in Table 2. 

Minimum. Maximum. 

Water, percent 8.02 11.27 

Ash, percent 1.33 3.06 

Protein, percent 9.19 . 16.75 

Fat, percent 2.22 5.36 

Fiber, percent 0.96 3.57 

Carbohydrates, percent 65.40 75-30 

Weight per 1,000 kernels, grams 13.60 39-40 

Weight per bushel, pounds 50.00 60.50 



8 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Table 2 also shows that the sorghum grains are very similar to each 
other in composition, the widest differences between the average com- 
position of different grains being as follows : 

Percent. 



Water 0.33 

Ash 0.31 

Protein 1.69 

Fat 1.05 

Fiber 0.29 

Carbohydrates 2.80 



In the case of the weight per 1,000 kernels, however, there is a 
material difference between the lowest and the highest average, 
namely, 14.7 grams, while the weight per bushel shows a difference 
of 2.6 pounds between the lowest and highest average. 

Table 3 shows the number of samples analyzed, the average weight 
per 1,000 kernels, and the average protein content of the various 
groups of grain sorghums for each year from 1908 to 1912, inclusive. 
Tables 4, 5, and 6 give comparative determinations for sorghum 
grains with low and high protein content, low and high weight per 
1,000 kernels, and low and high ash content, respectively. 



Table 3. — Average percentage of protein and weight per 1,000 kernels of the 
various grain sorghums grown at Amarillo, Tex., each year from igo8 to igi2. 



Crop. 


1908. 


1909. 


1910. 


No. of 
samples. 


Protein. 


Weight 
per 
1,000 
kernels. 


No. of 
samples. 


Protein. 


Weight 
per 
1,000 
kernels. 


No. of 
samples. 


Protein. 


Weight 
per 
1,000 
kernels. 


Milo 

Kaoliang .... 

Kafir 

Durra 

Durra kafir. . 


IS 

10 

22 
6 


Percent 
11.31 
12.81 
12.16 
12.48 


Grams 
34.80 
25.02 
24.00 
26.93 


16 
15 
38 
15 


Percent 

13-34 
13-66 
13.62 
14.03 


Grams 
36.00 
18.70 
22.23 
26.80 


22 
12 
32 

3 
12 


Percent 
13-43 
13.87 

13- 83 
14.00 

14- 48 


Grams 
34-35 
19-89 
18.40 
28.60 
21.56 





1911. 


1912. 


Crop. 


No. of 




Weight per 


No. of 




Weight per 




samples. 


Protein. 


1,000 kernels. 


samples. 


Protein. 


1,000 kernels. 






Percent 


Grams 




Percent 


Grams 


Milo 


45 


10.83 




52 


13-57 


33.38 


Kaoliang .... 


25 


11.88 


18.44 


30 


14.47 


17.86 


Kafir 


42 


12.49 


21.98 


48 


13.48 


18.96 


Durra 


6 


12.56 


29.10 


9 


14.19 


30.50 


Durra kafir. . 


19 


13-44 


22.07 


16 


14.88 


23.80 



LI'XLKRC & 1?AILKV: COMPOSITION OF GRAIN SORGHUMS. 9 

A study of liable 3 shows that in the grain sorghums there is no 
well defined relation between the weight per 1,000 kernels and the 
protein content, such as one might expect and such as usually exists 
in the case of wheat. In the milos it would seem that a high protein 
content is accompanied by a high weight per 1,000 kernels, while 
with the other sorghums there is a tendency for the opposite to be 
true, that is, for high protein samples to be those of low weight per 
1,000 kernels. In this respect these latter are like wheat, a high 
weight per 1,000 kernels being correlated with a low protein content. 
These relations are brought out more clearly in Table 4, where it may 
also be definitely seen that in most cases high protein samples also 
contain low fat, high fiber, and high ash. One reason why high 
weight per 1,000 kernels of milos and of durras is correlated with 
high protein content would be evident if the bran from these grains 
were found to contain a lower percentage of protein than the rest of 
the grain. As large kernels contain relatively less bran, they would 
be, therefore, in this case relatively richer in protein. In wheat the 
bran is richer in protein than the rest of the grain and, therefore, high 
protein wheat usually weighs less than low protein wheat. 



Table 4. — Comparison of grain sorghums of low and high protein content, 
showing average figures for other determinations. 



Crop and class. 


No. of 
samples. 


.S? 


Water. 


in 
< 




Fiber. 


Carbohy- 
drates. 


Weight per 
1,000 kernels. 


Weight per 
bushel. 






% 


% 


% 


% 


% 


% 


Gm. 


Lbs, 


Milo. 






















16 


10.69 


9-36 


1-59 


3.27 


1-43 


73-69 


34-9 


58.2 




18 


13-75 


9.25 


1.66 


3-08 


1-50 


70.76 


36.1 


58.1 


Dwarf milo. 




















Protein less than 9.62 percent 


14 


9-56 


9.48 


1-54 


3-37 


1-39 


74-65 


29.8 


58.8 


Protein more than 13. 150 percent 


14 


13-56 


9-57 


1.67 


3-23 


1.53 


70.48 


32.1 


57.7 


Brown kaoliang. 






















17 


11.44 


9-35 


1.84 


4.17 


1.27 


71-96 


19.0 


56.7 


Protein more than 14.70 percent 


18 


14-75 


9.88 


1-95 


4.07 


1.50 


67.94 


17.7 


55-2 


Blackhull kafir. 




















Protein less than 12.60 percent. ...... 


19 


12.56 


9-45 


1.66 


3-53 


1-58 


71-30 


23-1 


58-7 


Protein more than 15.20 percent 


19 


15-25 


9.60 


1-85 


3-44 


1.57 


68.22 


2.0.8 


58.1 


Red kafir. 






















15 


II. 13 


9.92 


1.60 


3-12 


1-44 


72.80 


21.5 


58-6 


Protein more than 13.70 percent 


16 


13.75 


9.82 


1-79 


3-22 


1.46 


69.96 


20.5 


57-9 


Durra and durra kafir. 




















Protein less than 12.46 percent 


10 


11.90 


9.55 


1.87 


3.57 


1.39 




24.4 


57.0 


Protein more than 14.60 percent 


13 


15.08 


9.08 


1-95 


3-58 


1.63 




23-7 


55-5 



That a large kernel is often correlated with a low fiber content is 
due to the fact that large kernels have relatively less superficial area 



lO JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

than small and consequently less bran, in which the greater part of 
the fiber is found. A grain with low weight per i,ooo kernels being 
smaller and having a relatively large surface or coating will be higher 
in fiber than one having a higher kernel weight. It may also be higher 
in' protein if the bran has a high protein content, as is true of most 
cereals, especially wheat, rye, oats, and barley. Low protein in the 
grain sorghums is, on the other hand, correlated with high weight per 
bushel, low fiber, low ash, and low pentosans, and, except in the case 
of milos, with low weight per i,ooo kernels.^ 

The above conclusions were drawn after comparing the average 
composition of a number of samples containing the least amount of 
protein with a number containing the greatest amount of protein. 



Table 5. — Comparison of grain sorghums of low and high weight per 1,000 
kernels, showing average figures for other determinations. 







per 
nels. 












1) 

p.—' 


Crop and class. 


No. c 

ampl 


eight 
)o ker 


Wate 


< 




fa 


<u 

XI 

fa 


eight 
bush< 




in 




















Gm. 


% 


% 


% 


% 


% 


Lbs. 


Milo. 


















Wt. of 1,000 kernels less than 30.2 grams . . 


12 


28.2 


9.16 


1.65 


11-33 


3-26 


I-S3 


57.7 


Wt. of 1,000 kernels more than 37.9 grams. 


II 


38.7 


9.64 


1.52 


11.90 


3.31 


1-51 


58.3 


Kaoliang. 


















Wt. of 1,000 kernels less than 16.6 grams . . 


II 


15.4 


8.90 


1.^5 


12.70 


4.18 


1.46 


55.8 


Wt. of 1,000 kernels more than 23.4 grams. 


II 


25.4 


8.94 


1.82 


12.90 


4.27 


1.29 


56.4 


Kafir. 


















Wt. of 1,000 kernels less than 16.9 grams . . 


13 


iS-S 


9.18 


1.70 


12.57 


3-25 


1.63 


57.2 


Wt. of 1,-000 kernels more than 26.3 grams. 


9 


28.1 


9.48 


1.65 


12.61 


3-36 


1.38 


58.4 


Durra kafir. 


















Wt. of 1,000 kernels less than 21.0 grams . . 


12 


19.7 


9.08 


1.92 


13-62 


3-64 


1.63 


57.1 


Wt. of 1,000 kernels more than 28.0 grams. 


10 


32.7 


9.40 


1.70 


13.22 


3-22 


1-45 


55-7 



From a consideration of the results given in Table 5 it would seem 
that samples having a high weight per 1,000 kernels are those which 
contain low fiber and low ash. This table also shows that there is no 
definite relation between the weight per 1,000 kernels and the protein 
content. From Table 6 it will be seen that low ash is accompanied 
by low protein, low fiber, high weight per 1,000 kernels, and high 
weight per bushel. There is, however, no such correlation between 
ash and fat content. 

5 The pentosans were not determined in all samples and consequently the 
results are not recorded here. The amount of pentosans in the samples 
analyzed for this constituent varies from 2.35 to 4.69, with an average of 3.89. 
High pentosan content is accompanied by high fiber, high ash, high protein, 
and low weight per 1,000 kernels in all the groups except the milos, where 
high pentosan samples have low fiber, low ash, low protein, and high weight 
per 1,000 kernels. 



LECLERC ."v: RAILEV COM I'OSITTON OF. GRAIN SORGHUMS. II 



Table 6. — Comparison of grain sorghums of loiv and high ash content, showing 
average figures for other determinations. 





















Crop and class. 


No. of 
sample: 


in 

<; 


.s ^ 

^> 


Water. 




Fiber. 


Weight p 
1,000 kern 


Weight I 
bushel 






% 


Of 


% 


/o 


% 


Cm. 


Lbs. 


TV T ; 1 « 

MllO. 




















10 


1.42 


10.91 


9-78 


3-32 


I-51 


35-1 


58.8 


Ash more than 1.77 percent 


12 


1. 81 


13-51 


8.66 


3-03 


1.68 


32.5 


57.2 


Kaoliang. 


















Ash less than 1.70 percent 


10 


1.62 


12.63 


8.95 


4-47 


1.26 


19.2 


57-5 


Ash more than 2.05 percent 


9 


2.13 


13-47 


8.80 


4-23 


1.40 


18.8 


55.2 


Kafir. 


















Ash less than i.S9 percent 


II 


1.55 


12.01 


9.72 


3-30 


1-49 


23.6 


58.S 


Ash more than 1.90 percent 


II 


1.94 


14.85 


9.36 


3-58 


1.6S 


20.7 


57.1 


Durra kafir. 




















II 


1.59 


12.60 


9.26 


3-45 


1.32 


29.1 


57-2 


Ash more than 2.00 percent 


12 


2.06 


14-55 


8.66 


3-62 


1-73 


22.0 


54-7 



Two of the less extensively grown sorghums are broomcorn and 
shallu. The figures given in Table 7 show the composition of a small 
number of samples of seed of these crops grown at Amarillo, Tex., 
during the years 191 1 and 1912. It will be noted that the samples 
from the crops grown in 191 1 were considerably lower in protein, ash, 
and fiber, but somewhat higher in fat and in carbohydrates than were 
those grown in 1912. 



Table 7. — Composition of broomcorn and shallu grown at Amarillo, Tex. 



Name. 


Water. 


Ash. 


Protein 
(A^X 
6.25). 


Fat. 


Fiber. 


Carbo- 

hy- 
drates. 


Weight 
per 
1,000 
kernels. 


Weight 

per 
bushel. 




Percent 


Percent 


Percent 


Percent 


Percent 


Percent 


Grams 


Pounds 


Broomcorn. 


















1911, 5 samples. 


















Minimum 


10.05 


2.56 


11.94 


3-43 


3-40 


65.66 


14.7 


50.3 


Maximum 


10.72 


3.02 


12.81 


3-85 


4-39 


67.82 


16.3 


53.9 


Average 


10.52 


2.85 


12.44 


3.66 


3-96 


66.54 


15-I 


51.6 


1912, 9 samples. 




















9-36 


. 2.84 


12.44 


2.95 


4.17 


62.24 


11.4 


42.6 




9-83 


3-6o 


15-94 


3.56 


6.58 


66.49 


16.4 


51.4 




9.62 


3-16 


14.63 


3-34 


5-54 


63.62 


13-9 


48.5 


Shallu. 


















1911, 4 samples. 




















10.74 


1.84 


12.88 


3-51 


1-65 


67.42 


14.0 


58.5 




10.98 


1.91 


14-13 


4-13 


1.80 


68.97 


17.0 


59-6 




10.85 


1.88 


13.31 


3.72 


1-74 


68.48 


14.9 


58.9 


1912, 6 samples. 


















Minimum 


9.68 


1.94 


16.13 


3-19 


1.86 


65-38 


14.7 


52.2 




10.76 


2-37 


16.69 


4.02 


2.62 


66.15 


15-9 


58.S 




10.07 


2.05 


16.44 


3-67 


2.05 


65-75 


15.3 


57.2 



12 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

It would seem from a study of Table 7 that the weight per 1,000 
kernels of shallu was not influenced either by the abundant rainfall 
at Amarillo during 191 1 (16.44 inches from April to ripening), or by 
the scanty rainfall in 1912 (10.68 inches). In the case of the broom- 
corn, however, a slight decrease in weight per 1,000 kernels in 1912 
may have been the result of a dry season. The protein content, how- 
ever, was materially affected by a lack of rainfall in that year. 

From a consideration of the data in Table 8, it is impossible to 
state whether the low protein in the grains of 1908 and 1911 was due 
to the high annual rainfall, to the high rainfall from April to the 
ripening period, or to the rainfall from April up to the period of emer- 
gence. Although the experiments discussed in this paper have been 
carried on for five years and with a large number of varieties of sor- 
ghums, the data obtained are not sufficient to reveal the reason for 
the low percentage of protein in the grains during the two years, 1908 
and 191 1. It is necessary, apparently, to obtain similar data not only 
for a number of years and with a number of crops, but in a number 
of different localities. 

Because in the years 1908 and 1911 many more pounds of protein 
were produced per acre than in the years 1909, 1910, and 1912, and 
because in 1908 and 191 1 there were very copious rains before the 
period of emergence, it may be concluded that this large amount of 
rainfall gave the plants a propitious start, allowing them to become 
well developed and capable of bearing large crops. During a season 
of high and favorable rainfall the yield is much increased, but the 
crop contains a relatively low percentage of protein. 

Table 8 gives the total rainfall from April to the ripening period, 
and also the rainfall from April to emergence, from emergence to 
heading, and from heading to ripening. This table also gives the 
yield per acre in bushels and in pounds, the percentage of protein, 
and the yield of protein per acre in pounds. The figures show that 
in 1908 and 191 1 the largest yields were obtained; that these crops 
contained the smallest percentage of protein; that notwithstanding 
the low percentage of protein the number of pounds of protein per 
acre was much greater than in the other three years ; that the greatest 
amount of rain fell in 1908 and 1911 ; and that the amount of rain 
faUing before the period of emergence was considerably larger in 
these years than in other years. 



LECLERC & bailey: COMPOSITION OF GRAIN SORGHUMS. 



•3 



Table 8. — Rainfall at various periods during the growth of various grain sor- 
ghums at Amarillo, Tex., annually from igoS to i<)i2, zmth the yield 
of grain and of protein per acre each year. 



Crop iind year. 


Pro- 
tein. 


Total, 
April 
to rip- 
ening.o 


April to 
emerjf- 
ence. 


Rainfall. 

F'mersence 
to heading. 


Heading to 
ripening. 


Yield of grain 
per acre. 


Yield of 

protein 
per 
acre. 




% 


Inches 


Inches 


Inches 


Inches 




Bu. 


Lbs. 


Lbs. 


Milo. 
























II. 31 


15.12 


5.35 


7.88 




2.75 




^■^.62 


2,053 


232.2 


1909 


13.44 


12.99 


1.56 




,8.58 




2.92 




6.14 


355 


47.7 




13-44 


10.00 


3.37 


\ 


' 3. II 1 
.5.58 J 


6 


2.96 




19.67 


1. 143 


153.6 


1911 


II. 13 


16.20 


8.55 


3-93 




3.22 




32.28 


1,890 


210.4 


1912 


13.63 


9.16 


2.17 


5.01 




1.76 




18.97 


1,107 


151.0 


Dwarf milo, 






















1908 


11.31 


15.12 


5.35 


7.88 




2.75 




41.37 


2,388 


270.0 


1909 


13,13 


12.99 


1.56 




8.58 




2.92 




11.00 


638 


83.8 


1910 


13.13 


10.00 


3-37 


\ 


^3.iil 
.5.58 J 


h 


2.96 




20.68 


1,190 


156.3 


1911 


10.13 


16.20 


8.55 


3-93 




3.22 




38.24 


2,240 


227.0 


1912 


13.31 


9.16 


2.17 


5.80 




1.76 




22.64 


1,318 


175-4 


White durra, 
























12.63 




6. II 




1 5. II 1 
I 8.65 1 


b 




r 5.83 1 
I 6.57 , 


c 


33.29 


1,870 


236.2 


1909 


14.06 


12.99 


1.76 




r8.23 1 

8.90 


b 




1 0.76 1 
2.92 


c 


II. 51 


647 


91.0 




14.00 


10.06 


3.46 




3. II ' 
I 4.62 


b 


2.96 




10.60 


602 


84.3 


I9II 


12.56 


16.62 


8.55 


3.83 




2.81 






1 ,840 


231.0 


I9I2 


14.19 


II.3I 


2.17 


\ 


1 4-97 1 
I 5.76 J 


1 


\ 


I 1.66 1 
13.81 j 


c 

\ 


19.17 


1,084 


153.3 


Brown kaoliang, 














f 2.75 ^ 










1908 


12.75 


17.16 


5-35 




7.88 




\ 


14.37 J 


\ 


29.71 


1,675 


213.5 


1909 


13.75 


11.47 


1.77 




^ 7.43 ^ 
1 8.92 , 






3.65 




13.04 


738 


101.5 




13.75 


10.00 


3.47 




2.27 1 
14.78 j 


r 




4.70 1 
I 5.33 J 


c 


10.45 


583 


80.2 




11.94 


10.01 


8.44 




3.93 






1 2.81 1 
3.22 ^ 


c 


22.09 


1,244 


148.5 




14.44 


11.24 


2.15 


\ 


r3.i2^ 
1 6.67 J 


r 




1 4.02 ' 

I 6.11 ; 


c 


11.70 


649 


93.6 


Blackhull kaoliang, 














( 3-49 ' 












12.44 


17.16 


5.35 


7.88 




< 


14-37 . 


y 


43.10 


2,370 


295.0 


1909 


13.25 


12.99 


1.77 


8.92 




2.82 




9.44 


544 


72.1 




13.81 


10.05 


3.47 


< 


1 3-95 


r 


2.96 




6.92 


384 


53.3 












16.07, 














1911 


11.25 


16.20 


8.44 


3.93 




3.22 




25.40 


1.465 


164.8 


1912 


14.75 


11.24 


2.15 


6.67 




2. IS 




15-93 


902 


132.7 



14 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Crop and year. 


Pro- 
tein. 


L otal, 
April 
to rip- 
ening.a 


April to 
emerg- 
ence. 


Rainfall. 

Emergence 
to heading. 


Heading 
to ripening. 


Yield of grain 
per acre. 


Yield of 
protein 
per 
acre. 




% 


Inches 


Inches 


Inches 


Inches 




Bu. 


Lbs. 


Lbs. 


Blackhull kafir, 






















12.63 


16.50 


5-i6 






2.72 








249-5 










\ 8.86 / 




I 2.95 J 










1909 


14-75 


12.99 


1.56 


9.01 


2.16 




5-04 


294 


43-3 




14.56 


10.07 


3-47 




1-45 




9.66 


555 


80.8 






















1911 


13-38 


16.62 


8-55 


6.62 


0.95 




21.24 


1,210 


162.0 




14.44 


11.31 


2.19 


6.27 


2.15 




4.09 


238 


34-3 


Red kafir, 




















1908 


11.31 


16.50 


5.26 


8,86 


3-10 




33-05 


1,960 


221.7 


1909 


12.50 


12.99 


1-56 


9.01 


2.16 




3-81 


224 


28.0 




12.31 


10.07 


3-47 


5-81 


0-35 




5-21 


300 


36.9 




11.56 


16.62 


8.55 


6.62 


0.95 




18.68 


1,090 


126.0 


1912 


12.94 


II.3I 


2.19 


6.27 


2.15 




4.26 


246 


31-8 


Average, 

1908 

1909 

1910 

1911 

1912 


12.05 
13-55 
13-57 
II. 71 
13-96 


16.24 

12.77 

10.04 

16.44 

10.68 


5.42 
1-65 
3-44 
8.30 
2.17 


/ 7-51 \^ 
I 8.27 [ 
/8.54\^ 
\ 8.84 f 

/ 3-74 r 

I 5-49 / 
4.68 

(5.44\^ 
\6.o6 / 




^ 3-34I 
i 3-81 J 
1 2.55 1 
1 2.80 ^ 
1 2.62 ' 
I 2-74 < 
1 2.45 1 
2.52 ^ 
1 2.23 
I 2.84 J 


c 
c 
c 
c 
c 


35-71 
8.57 
11.88 
27.19 
13.40 


2,041 

493 
680 
1,568 
795 


245-4 
66.8 
91. 1 
181.4 
1 10.3 



^ The total rainfall from harvest to harvest is approximately as follows : 
1907-08, 20.3 inches; 1908-09, 16.6 inches; 1909-10, 15.5 inches; 1910-11, 20.3 
inches; and 1911-12, 14.8 inches. 

* The heading v^as irregular. The top figures indicate the rainfall up to the 
first appearance of heads; the lower figures show the rainfall to the end of 
heading. 

The ripening was irregular. The top figures indicate the rainfall from 
beginning of heading to beginning of ripening. The lower figures show the 
rainfall from beginning of heading to end of ripening. 

From Table 8 it appears that the amount of rain falHng between 
the period of emergence and heading, or between the period of head- 
ing and ripening, did not exert so much influence upon the compo- 
sition or yield as did the early rainfall or the total rainfall. For ex- 
ample, in every case where there was a heavy early and total rainfall, 
the yield was very large and the percentage of protein was relatively 
small. On the other hand, for the rainfall between the periods of 
emergence and heading no such regularity is noted. For example, 
milo, dwarf milo, brown kaohang, blackhull kafir, and red kafir had a 



lecli:rc: ^S: i?Airj'\' : com i'osition oi- crain soiu'.ihims. 15 

high ])crccntai;c of protein in 1909, and yet the rainfall between the 
periods of emergence and heading was practically the same as in iQOcS 
when these same sorghums contained a low percentage of protein. 
In the same way milo, dwarf milo, white durra, and blackhull kao- 
liang in 191 1 contained low percentages of protein and in 1910 high 
percentages of protein, although in these years the total rainfall be- 
tween the periods of heading and ripening was almost identical. 
From the results obtained in these particular cases it is evident, there- 
fore, that the amount of rainfall between the periods of emergence 
and ripening has less influence upon the composition and yield of the 
crop than has either the amount of rainfall before the period of emer- 
gence or the total yearly rainfall. 

In Table 9 the average analyses of the grain sorghums are given. 
For comparison, average analyses of other grains and seeds grown 
generally throughout the United States are also given in this table. 



Table 9. — Average composition of the various grain sorghums at Amarillo, 
Tex., igo8-igi2, with the composition of various grains and seeds 
grown widely throughout the United States for comparison. 



Crop. 


No. of 
sam- 
ples. 


Water. 


Ash. 


Protein 
(A'X 
6.25). 


Fat. 


Fiber. 


Carbo- 

by- 
drates. 


Weight 
per 
1,000 
kernels. 


Weight 

per 
bushel. 






% 


% 


% 


% 


% 


% 


Gm. 


Lhs. 


Durra 


39 


9-50 


1-73 


13-63 


3.47 


1.49 


70.30 


28.2 


56.7 


Durra kafir 


47 


9.61 


1-95 


14.19 


3.51 


I.6S 


69.08 


22.3 


55-6 


Kafir 


182 


9.70 


1.76 


13-13 


3-30 


1.54 


70.30 


20.7 


58.2 


Kaoliang 


92 


9.37 


1.87 


13.37 


4.23 


1.36 


69.84 


19 2 


56.3 


Milo 


150 


9-39 


1.64 


12.50 


3.18 


1.52 


71.88 


33-9 


58.0 


Shallu 


10 


10.36 


1.98 


15.19 


3.69 


1.93 


66.84 


15. 1 


57.9 


Broomcorn 


14 


9-93 


3-05 


13.44 


3-45 


4-98 


64.66 


14.3 


49-6 


Maize** 


114 


10.04 


I-S5 


10.39 


5.20 


2.09 


70.69 


367.4 




Wheat^ 


166 


10.62 


1.82 


12.23 


1.77 


2.36 


71.18 


38.6 


60.4 


Oats 


133 


7.10 


3-50 


14-87 


4-52 


10.10 


59.91 


22.9 






84 


8.71 


2.98 


11.86 


' 2.02 


5-76 


68.98 


39.2 




Rye^ 


57 


8.67 


2.09 


11.32 


1.94 


1.46 


74.52 


20.7 




Emmer 


41 


8.50 


4.00 


13-94 


1-95 


8-44 


63-17 


29.6" 




Einkorn'' 


4 


8.34 


5-57 


14.67 


2.19 


13-55 


55.68 






Proso 


43 


9.12 


2.90 


13-54 


3-52 


7-38 


63-54 


5-3 




Millet 


13 


9.28 


3-47 


14.56 


4.17 


8.25 


60.27 


1.8 




Buckwheat 


31 


9-38 


2.05 


12.00 


2.47 


10. II 


63-99 


27.9 




Rice (hulled) 


97 


10.61 


1.25 


6.75 


2.35 


.85 


78.19 


23.7 




Soybeans 


33 


7.29 


5-19 


38.94 


18.14 


4.24 


26.20 


154.0 






16 


8.84 


3.79 


23.26 


1.25 


4-47 


58.39 


147-0 




Beans (adzuki) 


13 


10.01 


3-41 


2356 


.43 


4.11 


58.48 


87.0 




Flax 


64 


5-31 


3.67 


27.00, 


35.13 


5-09 


23.80 


4.0 





«U. S. Dept. Agr., Bur. Chem. Bui. 50. 1898. 
» U. S. Dept. Agr., Bur. Chem. Bui. 45. 1895. 
c Without hulls. 

<^U. S. Dept. Agr., Bur. Chem. Bui. 120, 1909. 



1 6 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



SUMMARY. 

The minimum and maximum results obtained and the average com- 
position found in analyses of several hundred samples of grain sor- 
ghums grown at Amarillo, Tex., from 1908 to 1912, are given. 

The results have been arranged in tabular form so as to bring out 
the correlations between the percentage of protein and the weight per 
1,000 kernels; and also to show how the composition of grain varies 
with 

( 1 ) a high and low protein content, 

(2) a high and low weight per 1,000 kernels, and 

(3) a high and low ash content. 

There is also an attempt made to correlate the protein content and 
the yield with the rainfall. 

Analyses of shallu and broomcorn are given and also of bread made 
in part from grain sorghum meal. 

The composition of the grain sorghums is compared with that of 
many other grains and seeds. 

LITERATURE CITED. 

1. Baird, R. O. 

1910. The chemistry of the Kafir corn kernel. Okla. Agr. Expt. Sta. 
Bui. 89, 15 p. 

2. , and Francis, C. K. 

1910. The chemical composition of Kafir corn. In Jour. Ind. Chem., 

V. 2, no. 12, p. 531. 

3. Ball, Carleton R. 

igio. The history and distribution of sorghum. U. S. Dept. Agr., Bur. 
Plant Indus. Bui. 175, 63 p., 15 fig. 

1911. The importance and improvement of the grain sorghums. U. S. 

Dept. Agr., Bur. Plant Indus. Bui. 203, 45 p., 13 fig. 

5. , and RoTHGEB, Benton E. 

• 1913. Kafir as a grain crop. U. S. Dept. Agr., Farmers' Bui. 552, 19 p., 
8 fig. 

6. . 

1915. Use of sorghum grain. U. S. Dept. Agr., Farmers' Bui. 686, 15 
p., 12 fig. 

7. Chamberlain, Joseph S. 

1908. Official and provisional methods of analysis. U. S. Dept. Agr., 
Bur. Chem. Bui. 107 (rev.), 272 p., 13 fig. 



Journal of the American Society of Agronomy. 



Plate I. 




Fig. I. Wheat grown in greenhouses kept at different temperatures; from left 
to right, grown at temperatures of 58. 62, 65, and 75° F., respectively. 




Fig. 2. Oats grown in greenhouses kept at different temperatures; from left 
to right, grown at temperatures of 58. 62. 65, and 75° F.. respectively. 



Journal of the American Society of Agronomy. 



Plate 2. 




Fig. I. Barley grown in greenhouses kept at different temperatures; from left 
to right, grown at temperatures of 58, 62, 65, and 75° F., resjiectively. 





Fig. 2. Rye grown in greenhouses kept at different temperatures ; from left 
to right, grown at temperatures of 58, 62. 65, and 75° F., respectively. 



IIUTCHESON & QUANTZ: CUOWING CICKKALS IN (iKlCI'.N 1 1 ()USi:S. I7 



THE EFFECT OF GREENHOUSE TEMPERATURES ON THE 
GROWTH OF SMALL GRAINS.^ 

T. B. HUTCIIESON AND K. E. QuANTZ. 

During the past few years it has been a common practice at some 
of the experiment stations to grow small grains in greenhouses for 
breeding work, for pot tests of soils, and for other similar experi- 
ments. For the crossing of varieties and the rapid increase of se- 
lected seed of small grains, the greenhouse offers almost invaluable 
aid, as by its use two crops of most of the grains may be grown in 
one year. Greenhouses are also usually more conveniently located 
than field plots and are therefore more accessible to the workers. 

When the crops are grown at the proper temperatures they set seed 
well and hand pollination may be practiced very successfully. It is 
not at all difficult to get 80 percent or more of all crosses made to 
set seed when the work is done carefully. For crossing it is usually 
best to grow the plants in small pots, as these may be placed in any 
convenient position for working and may be segregated easily after 
the cross is made, either by placing them in another apartment or by 
separating them from other plants by canvas partitions. 

Failures to get small grains to set seed properly in greenhouses are 
often reported, however. The writers believe that such failures are 
often due to the maintenance of improper temperatures in the houses. 
For this reason an experiment was conducted during the winter of 
191 5 to test the effect of temperature upon the growth of winter 
wheat, oats, barley, and rye grown in the greenhouse. 

Fulcaster wheat, Culberson oats. Union Winter barley, and Abruzzes 
rye were used in this experiment. The seeds of these grains were 
sown in 4-inch clay pots. The pots were divided into four groups. 
Each group consisted of four series of four pots each, one series of 
each of the grains under test. Each group was placed in a different 
greenhouse. The greenhouses were kept as near as possible to the 
following temperatures: House No. i, 75° F. ; house No. 2, 65° F. ; 
house No. 3, 62° F. ; and house No. 4, 58° F. Owing to the great 
variation of the outdoor temperature it was found impossible to keep 
the temperatures in the houses from fluctuating widely. This was 

1 Paper No. 3 from the Department of Agronomy, Virginia Polytechnic In- 
stitute and Agricultural Experiment Station. Received for publication October 
25, 1916. 



1 8 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



particularly true during April and May, when the days became longer 
and warmer. The curves (figure i) illustrating the average weekly 
temperature of the houses show this variation. It will be noticed 
that there was always a difference in the temperatures of the houses, 
though it was impossible under our conditions to keep the tempera- 
tures as regular as desired. 




Fig. I. Average weekly temperature of the four greenhouses from December 

27 to May 23. 

The seeds were sown and the pots were placed in their respective 
houses on December 21, 191 5. Several seeds were sown in each pot 
and all pots were thinned to two plants as soon as the plants were well 
up. Notes were taken on the growth as often as it was thought nec- 
essary. The data thus recorded are given in Table i. On May 27 
the experiment was discontinued on account of the extreme heat in 
the houses. The tillers recorded in Table i include only those that 
were still green on that date. The length of culms is the average 
length of all tillers which produced heads. The tillers which had not 
headed were not measured. 

There are some very striking features in the results of this experi- 
ment. In the first place, the time of heading, blooming, and ripening 
of the different grains varied considerably in the different houses. 
The order of maturity was sometimes almost reversed. Thus, oats 



HUTCIIF.SON .Kr QUANTZ I CUOWING CEREALS IN GREENHOUSES. I (J 

was the first i^raln to head in the hottest house, while in the coolest 
house the rye was first to head, the oats heading last. In the second 
warmest house the results were somewhat similar to those in the first 
house, but were slightly modified. Here also the oats headed before 
the rye. 



Table i. — Data on growth of wheat, oats, barley, and rye in greenhouses kept 
at various temperatures. 

WHEAT. 



Hqusc 
No. 


Tem- 
pera- 
ture 
(°F.)- 


Dates of . 


Number per 
plant. 


Average 
length, 
inches. 


Emergence. 


Heading. 


Blooming. 


Ripening. 


Tillers. 


Heads. 


Culms. 


Heads. 


I 
2 

3 
4 


75 
65 

62 

58 


"Dec. 28 

do. 

do. 
Dec. 31 


May 10-27 
May 15 
Apr. 26 
May 2 


May 13-27 
May 20 
Apr. 29 
May 3 


May 27 


8.75 
8.00 
5-37 
1.25 


0.87 
1.75 
3.00 
1. 12 


35.05 
31.48 
45.05 
36.74 


4.24 
3.28 
4.38 
3.68 


OATS. 


I 

2 

3 
4 


75 
65 

62 

58 


Dec. 29 

do. 

do. 
Dec. 31 


Apr. 17 

do. 
Apr. 25 
May I 


Apr. 26 

do. 
Apr. 29 

May 5 


May 24-27 

do. 
May 27 


9.00 
5.62 
3.37 
1.50 


4.62 

3-50 
2.00 
1.29 


30.46 7.13 
32.12 8.75 
34.54 8.00 
30.461 7.70 


BARLEY. 


I 

2 

3 
4 


75 
65 

62 

58 


Dec. 29 

do. 

do. 
Dec. 31 


May 25 
May 16 
May 2 


May 26 
May 17 
May 5 




48.25 
23.87 
8.00 
1-75 


1. 12 
1.75 


21.29 
23.42 


2.00 
2.38 


RYE. 


I 

2 

3 
4 


75 
65 

62 

58 


Dec. 28 

do. 

do. 
Dec. 31 


Apr. 25- 
May 27 
Apr. 29 
Apr. 18 
Apr. 17 


May 3-27 
May I 
Apr. 25 
May 2 


May 27 


29.25 
5-00 
5-37 
1. 00 


1. 12 
1.62 
3-00 
1. 00 


31.30 
40.65 
45-05 
60.31 


4.64 
4.25 
4.38 
4.88 



The wheat stood the warm temperature only a little better than 
the rye. It did not tiller as much and produced stronger heads, but 
otherwise there was little difference. In the cool houses the wheat 
did well — about like it does under ordinary field conditions. Each 
plant produced from one to several heads, all well filled, which ri- 
pened very uniformly. The pots of wheat from the houses kept at 
different temperatures are shown in Plate i, figure i. 

The oats appeared to be affected least of all by variations in tem- 
perature. The most notable effect was that in the warmer houses 



20 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

the plants grew more rapidly and the culms were weaker, necessitat- 
ing tying up the plants. The plants were taller in the warm houses 
and produced more leaves and stalks (Plate i, fig. 2). There was 
no harmful effect of a serious nature due to heat up to 80° F. The 
same fact was observed a few years ago at the Maryland station by 
the junior author. There the oats grew in a hot greenhouse to a 
height of 5 or 6 feet, while the wheat and barley did not head out. 

The barley showed most strikingly the difference between warm 
and cool temperatures. In the cool house the barley was almost ripe 
when the experiment was brought to a close and fully ripe three days 
later. In the warmest house it had produced only an enormous mass 
of tillers and there were no indications that it would ever head. Plate 
2, figure I, shows this condition clearly. In the second warmest house 
the barley did slightly better, but at the close of the experiment had 
made no start toward heading. In the second coolest house the bar- 
ley had not tillered excessively but it was considerably behind that in 
the coolest house in development. 

The rye evidently does best at the lower temperatures. It was best 
in the coldest house at the time the experiment was concluded. Each 
rye plant in this house produced one long, well-filled ear of grain, 
while in house No. 3 the rye produced more heads, but they were not 
nearly so well filled with grain. In houses No. i and No. 2 the rye 
did not do well at all. It seems to have wasted all of its energy in 
producing a great mass of tillers that crowded each other so that there 
was no chance for development. The heads were slender and showed 
little indication of grain production. The time of ripening was de- 
layed at least several weeks, and stretched out over a long period. 
The rye plants from the various houses are shown in Plate 2, 
figure 2. 

The results of this experiment show the effects of temperature in 
the following ways : 

1. Except in the case of oats, a cool temperature seems to produce 
earher maturity, while high temperature stimulates a rank growth of 
tillers and thus wastes energy needed for the formation of heads. 

2. Barley is affected by heat, while wheat and rye also suffer con- 
siderably, though not as much as barley. Oats show very little ill 
effect from high temperature. This would indicate that they are 
better for soil work in greenhouses where temperature cannot be 
controlled closely. 

3. The setting of seed is best in the cooler houses, except with oats, 
which show no difference in this respect. 



salmon: ticm iM:KA'rrRi:s and ( i:ki;al i-uodcc tion. 21 

4. The tillering was greatest in the warmer temperatures and least 
in the cool houses, but the difference in the number of heads is over- 
come by the large percentage of seed produced per head, except 
with oats. 

5. It is advisable to grow grain at a moderately cool temperature in 
the greenhouse. From 55° to 70° F. seems to be about the proper 
temperature. 

Virginia Polytechnic Institute, 
Blacksburg, Va, 



THE RELATION OF WINTER TEMPERATURE TO THE DISTRI- 
BUTION OF WINTER AND SPRING GRAINS IN THE 
UNITED STATES.^ 

S. C. Salmon. 

The successful culture of winter grains in northern latitudes de- 
pends largely on winter temperature. As, however, many other fac- 
tors may determine whether a farmer sows winter or spring grain 
where both may be successfully grown, the distribution of v/inter 
cereals in relation to temperature is of general interest. The census 
reports and weather bureau records of the United States and Canada 
afford excellent material for studying this relation. Such a study was 
begun during the winter of 1913-14 and has since been continued at 
intervals. The maps (figs. 2 and 3) show the results obtained 
with winter and spring wheat. The distribution in the United States 
was plotted by counties on the basis of the census returns for 1909. 
The distribution in Canada is based largely on the cereal maps pub- 
lished by the Canadian Department of the Interior in 1909. The 
acreage for Canada is for 191 1. The isotherms connect points of 
equal daily minimum temperature for January and February, the 
coldest months of the year. 

The isotherm of 10° F. d^ily minimum temperature coincides re- 
markably well with the northern boundary of winter wheat culture if 
this boundary is taken as the line beyond which spring wheat is grown 
more commonly than winter wheat. The correlation is so close, in 
fact, that in general the isotherm divides the winter from the spring 
wheat belt. There are some exceptions to this statement, but it is 

1 Contribution from the Kansas Agricultural Experiment Station. Received 
for publication October 18, 1916. 



22 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



perhaps significant that in nearly every case spring wheat is grown 
south of the isotherm rather than winter wheat north of it. This 
would be expected, since it is possible to grow spring grain wherever 
winter grain is a success, but the converse is not true. 

There are several reasons why spring wheat is sometimes grown 
where winter wheat is produced successfully. In Washington, for 
example, late harvest of spring oats and barley prevents timely prepa- 
ration of the ground for winter wheat. Many farmers who are grow- 
ing wheat exclusively find that they can sow and care for a larger 
acreage if they grow both winter and spring varieties than if they 
grow either exclusively. Spring wheat is sometimes preferred to 
winter wheat where corn is the main crop, since it is often difficult 
to get a good stand of winter wheat after corn. In western Kansas 
and eastern Colorado the lack of sufificient fall rain to insure germina- 
tion often prevents the growing of winter wheat where it would 
otherwise be successful. 




Fig. 2. Distribution of winter wheat in the United States and Canada. Each 
large dot represents 100,000 acres and each small dot, 5,000 acres. The iso- 
thermal lines connect points of equal mean daily temperature during January 
and February. 



SAL^roN : Ti:ArrKRATURi:s and ckrical production. 23 

The change from spring to winter wheat when it is fully demon- 
strated that the latter is best is generally rather slow, due to con- 
servative markets and prejudice on the part of grower and buyer. 
For many years hard winter wheat sold at a discount as compared 
with either spring wheat or soft winter wheat. In certain localities 
west of the Rocky Mountains at the present time the hard winter 
varieties, Turkey and Kharkof, sell at a discount. 

Small acreages of winter wheat are grown in central and northern 
Minnesota and Wisconsin and in southeastern South Dakota. Special 
methods of culture, as seeding in corn stalks, or special conditions, as 
a heavy covering of snow, usually are necessary for its successful 
production. 




Fig. 3. Distribution of spring wheat in the United States and Canada. The 
circles represent 100,000 acres each and the dots, 5,000 acres. Isothermal lines 
connect points of equal mean daily minimum temperatures for January and 
February, 



The isotherms of 20° and 30° coincide very closely with the north- 
ern limits for winter barley and winter oats respectively, so far as our 
knowledge of the climatic limits of these crops allow conclusions. 



24 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Winter barley, according to Derr,^ is confined to the states south of 
the Ohio and Platte rivers and west of the Rocky Mountains. At 
Manhattan, Kans., winter barley survives sufficiently to produce a 
good crop about four years out of five. Stephens states that it is 
grown extensively in Oregon only in Wasco and Umatilla counties 
in the Columbia River basin.^ The data are not sufficient to show the 
northern Hmit of this crpp, but it apparently bears a close relation to 
the daily minimum winter temperature as shown by the isotherm 
of 20° F. 

Warburton* shows by means of a shaded map the area in the United 
States adapted to winter oats as indicated by general observation and 
by experiments. He recommends them for as far north as southern 
Maryland, central Tennessee, central Arkansas, and southern Okla- 
homa, and states that they are grown to a limited extent in Utah, 
Oregon, and Washington. There is almost perfect agreement be- 
tween the northern limit of the culture of this cereal in the eastern 
United States, as shown by this map, and the isotherms of 30° F., as 
shown in figures 2 and 3. 

The absence of any correlation between the northern limit of winter 
cereal culture and snowfall is surprising. In the eastern United 
States, for example, where the snowfall is much greater than in cen- 
tral Nebraska, Montana, or southern Alberta, winter wheat appears 
not to be able to survive as low temperatures as in the last-named 
locaHties. In fact, considerable spring wheat is grown in Ontario 
south of the isotherm of 10° F. The lack of any correlation may be 
because a heavy snow increases the moisture content of the soil in the 
spring and so increases the danger from heaving. A heavy snow in 
the spring might therefore overbalance the protection afforded by a 
snow early in the winter. 

Kansas Agricultural Experiment Station, 
Manhattan, Kans. 

2Derr, H. B. Winter barley. U. S. Dept. Agr. Farmers Bui. 518, p. 7. 
1912. 

s Stephens, D. E. Report of the Eastern Oregon Dry-Farming Branch Ex- 
periment Station, Moro, Oregon, 1913-14, p. 19. Oreg. Agr. Coll. Expt. Sta., 
1914. 

4 Warburton, C. W. Winter oats for the South. U. S. Dept. Agr. Farmers 
Bui. 436, p. 13- iQii- 



SKINNKR iS: iu:attie: fkrtiijzI'.ks and r.iMi-: i<i:nri ui;.mknt. 25 



INFLUENCE OF FERTILIZERS AND SOIL AMENDMENTS ON 

SOIL ACIDITY.! 

J. J. Skinner and J. H. Beattie. 

INTRODUCTION. 

An experiment with manures, fertilizers, and soil amendments con- 
ducted for eight years at the Arlington Farm of the U. S. Department 
of Agriculture has given an opportunity to study soil acidity, soil 
oxidation, and other biochemical factors as influenced by fertilizers. 

Seven crops were grown, wheat, rye, clover, timothy, corn, cow- 
peas, and potatoes. Continuous cultivation of the same crop on the 
same plot year after year was practiced. Each of these crops occu- 
pied an area of i square rod, making seven crops for each fertilizer 
treatment. Adjoining each treated plot was an untreated plot grow- 
ing the same crop. This served as a check or control for the fertil- 
ized plot, and in this paper the data of a treated plot are always com- 
pared with those of the- adjoining untreated plot. The experiment 
occupied portions of five sections of the farm, the soil varying con- 
siderably in the different sections, so no comparisons between widely 
separated checks should be attempted. In each case the only fair 
comparison that can be made is between the treated plot and the ad- 
joining untreated plot. 

The treated and untreated plots are divided by a 3-foot path, and 
the portions of the plot growing different crops are divided by 2.5- 
foot paths. The plots then consist of a strip i rod wide, with seven 
subdivisions each i rod square, and parallel to them was a like strip 
with similar subdivisions, used as a control, growing the same crop. 

The soil on which the test was made is a heavy silty clay loam, 
rather poor physically, and low in organic matter. It is an acid soil 
which, when limed in sufficient amounts to neutralize the surface 6 
inches, within a few months becomes acid again. 

The experiment was begun in 1907; clover, timothy, corn, cowpeas, 
and potatoes were planted in the spring, and wheat and rye in the 
fall. Each succeeding year each crop was grown on the same plot 
as in the previous year. The fertilizers and materials added were 
applied annually before planting. 

^ Contribution from the Office of Soil Fertility Investigations, Bureau of 
Plant Industry, U. S. Department of Agriculture. Published by permission of 
the Secretary of Agriculture. Received for publication July 28, 1916. 



26 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Samples were taken of the plots in the summer of 191 2, when the 
experiment had been running five years. Five borings were taken 
with a soil auger from each plot and a composite made. The lime 
requirements of the soil were determined by the Veitch method.^ 
The oxidation power of the soil and other studies were made, but 
this paper concerns itself with only the acidity or lime requirements 
of the differently treated plots. 

EFFECT OF SULFATES ON SOIL ACIDITY. 

Among the fertilizers used in the experiments were several sulfates, 
and it is interesting to note the effect of this class of compounds on 
the reaction of the soil when applied year after year. Those included 
in the experiment are calcium sulfate, ferrous sulfate, manganese 
sulfate, copper sulfate, and potassium sulfate. As mentioned, seven 
plots, each growing a different crop, were fertilized with each of the 
chemicals and ea,ch treated plot lay next to an untreated plot growing 
the same crop, which is used as a check. 

Calcium Sulfate. — Calcium sulfate was applied to the plots each 
year at the rate of 500 pounds per acre. The first application was 
made in 1907. In the spring of each year, the plots growing corn, 
cowpeas, and potatoes were treated and the fertihzer harrowed in 
shortly before planting time. The plots growing wheat and rye re- 
ceived their treatment each year in September, before the seeding 
was done. The clover and timothy plots were treated before seeding 
when the experiment was started in 1907 ; thereafter the calcium sul- 
fate was applied to the surface each year in the early spring. 



Table i. — Effect of calcium sulfate on the acidity of the soil producing various 
crops, as indicated hy the pounds per acre of CaCOs required to 
neutralise the surface 6 inches. 



Treatment. 


Crop. 


Wheat. 


Rye. 


Clover. 


Timo- 
thy.. 


Corn. 


Cow- 
peas. 


Pota- 
toes. 


Aver- 
age. 




2,130 
2,170 


1,710 
2,140 


1.430 
2,140 


1.780 
2,140 


1,700 
1,780 


1.780 
2,170 


1,420 
1,800 


1,707 
2,050 



Soil samples were taken from each individual plot in June, 1912. 
Five borings were made in each plot with a soil auger to a depth of 

2 Veitch, F. P. The estimation of soil acidity and the lime requirements of 
soils. Jour. Amer. Chem. Soc, 24: 1120-1128. 1902. 

. Comparison of methods for the estimation of soil acidity. Jour. Amer. 

Cliem. Soc, 26: 637-662. 1904. 



SKINNER & BKATTIE: FKRTl I.I Z i:US AND LIMIC RKOUIKKMKNT. 2/ 

6 inches, and a composite of the five borings made. These samples 
were used to make the Hme requirement tests. The lime requirement 
for each of the untreated plots and each plot receiving calcium sul- 
fate is given in Table i. The figures given are the number of pounds 
of calciimi carbonate required to neutralize an acre 6 inches as de- 
termined by the Veitch method. 

In each plot the soil to which calcium sulfate had been added is 
more acid than the untreated plot growing the same crop. The lime 
requirement of all the plots, untreated as well as those treated with 
calcium sulfate, is high, ranging from 1,400 pounds to 2,100 pounds 
per acre. The differences in lime requirement of the untreated and 
calcium sulfate treated plots in some cases is very small, yet in 
others the differences are very marked. For instance, the untreated 
wheat plot requires 2,130 pounds of lime and the calcium sulfate plot 
2,170 pounds, while with the clover plot the untreated soil requires 
1,430 pounds and the calcium sulfate plot 2,140 pounds. However, 
in each case the calcium sulfate plot has a higher lime requirement 
than its untreated check. The averages given in the last column 
show the untreated plots to have a lime requirement of 1,707 pounds 
and the calcium sulfate plots 2,050 pounds. 

Ferrous Sulfate. — Ferrous sulfate was used in the experiment at 
the rate of 50 pounds per acre. The wheat and rye plots received 
their application in the fall of each year, and the clover, timothy, corn, 
cowpeas, and potatoes in the spring. The effect of ferrous sulfa;te 
on the acidity of the soil treated for five years is quite marked. The 
fertilization was begun, in 1907 and the samples taken for analysis in 
1912. The results are given in Table 2. 



Table 2. — Effect of ferrous sulfate on the acidity of the soil producing various 
crops, as indicated by the pounds per acre of CaCO^ required to 
neutralize the surface 6 inches. 



Treatment. 


Crop. 


Wheat. 


Rye. 


Clover. 


Tim- 
othy. 


Corn. 


Cow- 
peas. 


Pota- 
toes. 


Aver- 
age. 


Soil + ferrous sulfate 


2,492 
2,492 


2,136 
2,848 


1,780 
2,136 


2,136 
2,492 


2,136 
2,848 


2,311 
3.292 


I.88S 
3.197 


2,125 
2,758 



An examination of the table shows that in all except one case the 
lime requirement of the soil is higher in the ferrous sulfate plots 
than in the untreated plots. Both the untreated and ferrous sulfate 
plots growing wheat required 2,492 pounds of lime. The differences 
in the lime requirement of the treated and untreated soil with the 



28 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

ether crops is very marked. The greatest variation occurred in the 
potato plot; here the untreated soil had a Hme requirement of 1,885 
pounds and the ferrous sulfate treated plot a lime requirement of 
3,197 pounds per acre. The averages given in the last column show 
an average Hme requirement for the seven untreated plots of 2,125 
pounds per acre, while the average for the seven ferrous sulfate 
treated plots is 2,758 pounds per acre. 

Copper Sulfate. — The details of the experiment with copper sulfate 
and of the other sulfates to be given are the same as those of the 
two sulfates already given. Copper sulfate was applied annually at 
the rate of 50 pounds per acre. The soil when examined was found 
to be much more acid in the copper sulfate plots than in the untreated 
plots. The results are given in Table 3. 



Table 3. — Effect of copper sulfate on the acidity of the soil producing various 
crops, as indicated by the pounds per acre of CaCOz required to 
neutralize the surface 6 inches. 



Treatment. 


Crop. 


Wheat. 


'Rye. 


Clover. 


Timo- 
thy. 


Corn. 


Cow- 
peas. 


Pota- 
toes. 


Aver- 
age. 


Soil untreated 

Soil + copper sulfate 


1,770 
2,136 


2,136 
2,492 


1,885 
2,136 


1,780 
2,848 


1,426 
3.026 


1.534 
1.780 


1,742 
1,780 


1.753 
2,314- 



The difference in acidity of the treated and untreated plots is very 
marked, except in the plots growing potatoes. In this case the acidity 
is only sHghtly more in the copper sulfate plot than in the untreated 
one. The average lime requirements of the seven plots receiving the 
chemical was 2,314 pounds per acre against i,753 pounds as an aver- 
age for the seven untreated plots. 

Manganese Sulfate. — The results with manganese sulfate show that 
it has the same effect on the reaction of the soil as the other sulfate 
compounds. Manganese sulfate was added annually at the rate of 
50 pounds per acre, and after five years of treatment the soil has be- 
come more acid than the unfertilized soil. In each case the treated 
plot is more acid than its check, except with the wheat plots, and here 
the lime requirements are the same. The average lime requirement 
of the seven manganese sulfate plots was 2,231 pounds per acre and 
that of the untreated plots, 1,963 pounds. The results for each plot 
are given in Table 4. 



SKiNNKK \ iu:attii:: i'i;R'ni.izi:KS am) i.imi', Ki-.nniKicMicNT. 29 



Table 4. — Illicit of )in}>i(jatu\u' sulfiilc on the ocidily of the soil f/roimurj various 
crops, as indicated by the pounds per acre of CaCOa required to 
neutralise the surface 6 inches. 











Crop. 








Treatment. 


Wheat. 


Rye. 


Clover. 


Tim- 
othy. 


Corn. 


Cow- 
peas. 


Pota- 
toes. 


Aver- 
age. 




1,780 
1,780 


2,136 
2,492 


1,426 
1,780 


1,742 
2.136 


1,780 
J.492 


2,136 
2,492 


2,743 
2,451 


1.963 
2,231 


Soil + manganese sulfate . . 



Potassium Sulfate. — The results of the experiment with potassium 
sulfate are given in Table 7 in another connection, but for the sake 
of continuity will be given here briefly in connection with the action 
of the other sulfates. The potassium sulfate was applied annually 
at the rate of 100 pounds per acre and, like the other sulfates, pro- 
duced a more acid soil than the untreated, which is true with each 
set of plots. The increase in lime requirement over the check with 
the wheat plots was 510 pounds per acre; the rye plots, 254 pounds; 
clover plots, 612 pounds; timothy plots, 1,502 pounds; corn plots, 
534 pounds; cowpea plots, 356 pounds; and the potato plots, 303 
pounds per acre. 

Potassium Sulfide. — Potassium sulfide was also used and its effect 
seems to be similar to that of the sulfate. Potassium sulfide was ap- 
plied annually at the rate of 200 pounds per acre. The lime require- 
ments of the different plots are given in Table 5. 



Table 5. — Effect of potassium sulfide on the acidity of the soil growing various 
crops, as indicated by the pounds per acre of CaCO^. required to 
neutralize the surface 6 inches. 





Crop. 


Treatment. 












Cow- 
peas. 






Wheat. 


Rye. 


Clover. 


Timo- 
thy. 


Corn. 


Pota- 
toes. 


Aver- 
age. 




2,848 


2,136 


2,136 


2,140 


1,780 


2,670 


2,090 


2,257 


Soil + potassium sulfide . . . 


2,918 


2,492 


2,492 


2,140 


2,140 


2,490 


2,140 


2,402 



The differences between the treated plot and its check in each case 
are small, yet in each case except two the lime requirement of the 
potassium sulfide treated plot is greater than that of the untreated 
plot growing the same crop. In the wheat plots the untreated soil 
had a lime requirement greater than the treated plot and in the timo- 
thy plots the lime requirement was the same in both. The average 
lime requirement of seven plots receiving potassium sulfide was 2,402 
pounds per acre, while that of the check plots was 2,257 per acre. 



30 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

The increased acidity of the soil in cases where the sulfates were 
added was probably due in some cases to the utihzation of the base 
by plants, as potassium and calcium, or by chemical reactions going 
on in the soil. Some of these compounds, as copper, manganese, and 
iron sulfate, have a catalytic action in the soil, affecting its oxidation, 
enzymotic action, and other life activities in such a way as to produce 
different effects in the soil, depending on its characteristics. With the 
soil in this experiment the conditions have undoubtedly been such as 
to promote the activities which tend to produce acidity ; in this case 
the effect of such compounds on the acidity is an indirect one and 
their influence would be different with different soils. The amount 
of copper, manganese, and iron sulfate added is not commensurate 
chemically with the acidity which has developed in the soil with its use. 

EFFECT OF CARBONATES ON SOIL ACIDITY. 

Calcium Carbonate. — In this soil amendment experiment lime 
(CaCOg) was used on several plots and in different amounts. One 
of these sets of seven plots was limed annually at the rate of 300 
pounds per acre, and while this soil did not remain neutral under the 
crop management practiced, its lime requirement at the end of five 
years was much less than its untreated or check plot. On other plots 
where lime was applied annually for five years at the rate of one ton 
per acre, the soil was found to be neutral. The soil under study is 
naturally an acid one and requires lime annually in considerable quan- 
tities to keep it in a neutral condition. 

Magnesium Carbonate. — In addition to the calcium carbonate in- 
teresting results were also secured with magnesium carbonate, which 
was appHed annually at the rate of 200 pounds per acre. At the end 
of five years the soil was found to be acid but not so much as the un- 
treated plots. In four cases the lime requirement was the same in 
the magnesium carbonate plot as its check plot, and in three cases the 
acidity was less where magnesium carbonate was used. The results 
of the lime requirement determinations are given in Table 6. 



Table 6. — Effect of magnesium carbonate on the acidity of the soil growing 
various crops, as indicated by the pounds of CaCO^ per acre required 
to neutralize the surface 6 inches. 



Treatment. 


Crop. 


Wheat. 


Rye. 


Clover. 


Timo- 
thy. 


Corn. 


Cow- 
peas. 


Pota- 
toes. 


Aver- 
age. 




1,780 
1,780 


2,136 
2,136 


1,426 
1,426 


2,136 
1,742 


1,780 
1,780 


2,136 
1,397 


2,743 
1,742 


2,018 
1,715 


Soil + magnesium carbonate 



SKINNER & HKATTli:: FKRT I IJ Z I:rS AND LIME REQUIREMENT. 3 I 

The average lime requirenietit of the seven treated plots was 1,7^5 
pounds per acre, and of the seven check plots 2,018 pounds per acre. 
In no case was the magnesium carhonate plot more acid than the 
check. 

EFFECT OF ACID PHOSPHATE, SODIUM NITRATE, AND POTASSIUM SULFATE ON 

SOIL' ACIDITY. 

Acid phosphate, sodium nitrate, and potassium sulfate, the com- 
monly used fertilizers, have different effects on the reaction of the 
soil. It is commonly accepted that soil continually fertilized with 
sodium nitrate, the nitrate being utilized, eventually becomes alkaline 
and that this alkalinity causes a sticky physical effect, producing an 
undesirable soil for crop growth and cultivation. Potassium sulfate, 
on the other hand, is considered to have the opposite effect, as the 
potassium is utilized, leaving the acid radical and tending to produce 
an acid effect in the soil. It has recently been shown by Conner^ at 
the Purdue University Agricultural Experiment Station that soils 
treated with acid phosphate for twenty years show less acidity than 
soils that have not had phosphate. This, however, probably varies 
with the nature of the soil, for in this experiment on the Arlington 
silty clay loam the acid phosphate plots became more acid than the 
untreated soil but not so acid as the potassium sulfate plot. 



Tarle 7. — Effect of fertilizers on the acidity of the soil growing various crops, 
as indicated by the pounds per acre of CaCOz required to neutralize 
the surface 6 inches. 





Crop. 


Treatment. 














Pota- 
toes. 






Wheat. 


Rye. 


Clover. 


Tim- 
othy. 


Corn. 


Cow- 
peas. 


Aver- 
age. 


Soil untreated, check A. . . . 


1,836 


1,607 


1,426 


1.780 


1,707 


1,780 


1,815 


1,708 


Soil + sodium nitrate 


1,636 


1,780 


1,280 


1,426 


1,426 


1,780 


1,780 


1,587 


Soil + acid phosphate 


2,136 


1,780 


1,780 


1.993 


1. 815 


2,171 


2,136 


1,973 


Soil untreated, check B. . . , 


1,726 


1,526 


1.346 


1,602 


1,602 


1,780 


1,938 


1,646 


Soil + potassium sulfate . . . 


2,136 


1,780 


I.9S8 


3.104 


2,136 


2,136 


2,241 


2,212 


Average of Checks A and B 


1,782 


1,566 


1.386 


1,691 


1.654 


1,780 


1,876 


1,677 



The plots with these fertilizer treatments lay adjoining; there were 
seven plots of each fertilizer treatment, each plot growing a different 
crop. An untreated set of plots lay next to each fertilized plot. The 
plots were situated so that the results from the differently treated 
plots growing the same crop were comparable with one another as 
well as with their check plots. The test was conducted in detail as 

* Conner, S. D. Acid soils and the effect of acid phosphate and other fer- 
tilizers upon them. Jour. Ind. Eng. Chem., 8: 35-41. 1916. 



32 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

those already given; acid phosphate was applied annually before 
planting at the rate of 200 pounds per acre, sodium nitrate at the rate 
of 150 pounds per acre, and potassium sulfate at the rate of 100 
pounds per acre. The results of the Hme requirement test made after 
five years of fertilization are given in Table 7. 

In the table the lime requirement of the two check plots is given. 
Check plot A lay first and next to the sodium nitrate plot, and check 
B is the fourth plot and lay between the acid phosphate and the potas- 
sium sulfate plot, serving as a check for both. It is observed that the 
lime requirement of the two check plots is very close together, check 
A being in most cases somewhat more acid. The averages of the two 
checks are given in the bottom line of the table and will be used in 
comparing the acidity of the fertihzed plots. 

Considering the sodium nitrate plots, all are shown to be quite acid, 
but only two show a higher lime requirement than their check. With 
the rye and potato plots the check plots are less acid, and in the cow- 
pea plots the lime requirement was the same in both treated and un- 
treated soil. In the wheat, clover, timothy, and corn plots the sodium 
nitrate soil was less acid than the untreated soil. The average lime 
requirement of the seven sodium nitrate plots was. 1,587 pounds per 
acre, against an average of 1,677 pounds for the checks. Where 
acid phosphate was used the acidity of the soil was found to be greater 
than in the unfertilized check. This was true in each of the plots. 
The average Hme requirement of the acid phosphate plots was 1,973 
pounds per acre and that of the check plot lying next to it was 1,646 
pounds per acre. The potassium sulfate plots were also in every case 
more acid than the checks. The average lime requirement of the 
seven potassium sulfate plots was 2,212 pounds per acre against an 
average of 1,677 pounds for the checks. The potassium sulfate 
caused a greater acidity than the acid phosphate. 

EFFECT OF STARCH AND MANURE ON SOIL ACIDITY. 

Among another class of substances in this experiment were several 
organic materials, and the acidity of the soil as affected by starch was 
determined. To the plots concerned potato starch was added an- 
nually at the rate of 500 pounds per acre. This was applied before 
planting and harrowed in, except on the clover and timothy plot, 
where it was spread on the surface in the spring of each year. 

Each of the seven starch plots were more acid than the correspond- 
ing check plots after five years of treatment. The differences in the 
lime requirements in the various plots are small, yet they are con- 
sistent throughout the series. The results are given in Table 8. 



SKINNER & BKATTIE: FKKTILIZKKS AND LIME KEOUJ REM ENT. 33 



Table 8. — Effect of starch on the acidity of the soil gronAng various crops, as 
determined by the pounds per acre of CaCOa required to neutralize 
the surface 6 inches. 



Treatment. 


Crop. 


Wheat. 


Rye. 


Clover. 


Timo- 
thy. 


Corn, 


Cow- 
peas. 


Pota- 
toes. 




Aver- 
age. 




2,850 
2,490 


2,490 
2,670 


2,140 
2,240 


2,040 
2,140 


2,240 
2,850 


2,530 
3^40 


2.530 
2,880 


2,403 
2,630 





As seen in the table, the average lime requirement of the seven 
starch plots was 2,630 pounds per acre, while that for the seven check 
plots was 2,403 pounds per acre. The decomposition of this organic 
matter in the soil has produced an acidity greater than is normal for 
this soil. The influence of organic materials incorporated in soils on 
its acidity is undoubtedly determined by the process of its decompo- 
sition, which is influenced by the nature of the material added, the 
character of the hfe processes in the soil, such as bacteria, enzymes, 
etc., and other soil characteristics. Materials of this nature may pro- 
duce an altogether different effect in some other soil which has better 
oxidative power and enzymotic activities. 

In this connection the effect of stable manure in producing acidity 
in the soil is interesting. Well-rotted stable manure was added to the 
soil in amounts of 4 tons per acre annually. To another plot the 
manure thoroughly leached was added, and to a third the leachings 
from the manure were added. The operation was done by spreading 
cheese cloth over the plots which were to receive the leachings only 
and spreading the desired amount of manure over the cloth. The 
boundaries of the plot were banked in such a way that none of the 
leachings could run away. This operation was done early in the 
spring, during a rainy spell, and additional water was sprinkled over 
the manure from time to time so as to assure a thorough and effective 
leaching. The manure after leaching consisted principally of the 
straw bedding material and coarse material of the manure. This was 
then put on the plot which was to receive the leached manure and 
worked in. Four tons of manure per acre were used for the leaching 
experiment. 

The experiment consisted of five plots, each divided into seven 
parts and each part growing a different crop. Plot No. i was a check 
and received nothing ; No. 2 received the stable manure ; No. 3 the 
leached manure ; No. 4 was a check ; and No. 5 received the leachings 
from the manure. These treatments were applied each year before 
planting, beginning in 1907. The lime requirement was determined 
in each of the plots in 1912 and is given in Table 9. 



34 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Table 9. — Effect of manure on the acidity of the soil growing various crops, as 
determined by the pounds per acre of CaCOa required to neutralise 
the surface 6 inches. 



Plot 
No. 


Treatment. 


Crop. 


Wheat. 


Rye. 


Clover. 


Timo- 
thy. 


Corn. 


Cow- 
peas. 


Pota- 
toes. 


Aver- 
age. 


I 


Soil untreated 


2,367 


2.476 


2,536 


2,436 


2,376 


2,492 


2,626 


2,473 


2 


Soil + manure .... 


2,848 


2,492 


2,562 


1,780 


2,492 


1,426 


1,780 


2,197 


3 


Soil + leached 






















3.282 


3.560 


3,104 


2,492 


2,492 


3.104 


3.560 


3,085 


4 


Soil untreated 


2,490 


2,613 


2,753 


2,692 


2,431 


2.543 


2,451 


2,567 


5 


Soil + leachings . . . 


1.958 


2,136 


2,136 


2,136 


1,426 


2,241 


2,136 


2,024 




Average of checks . 


2,428 


2,544 


2,644 


2,561 


2,403 


2.517 


2,538 


2,520 



The table shows that as an average of seven manured plots the lime 
requirement is less in the manured plot (No. 2) than in the check 
plots (Nos. I and 4). With four of the crops the manured plot was 
less acid than its check ; with the other three crops the manured plots 
were more acid than the checks growing the same crop. The plot 
receiving the leached manure (No. 3) was much more acid than its 
check or the manured plot (No. 2), and the plot receiving the leach- 
ings was not as acid as the checks or any of the manured plots. 

Considering the averages of the seven plots of each treatment, 
check No. i had a lime requirement of 2,473 pounds per acre; the 
second check (plot No. 4) had a Hme requirement of 2,567 pounds 
per acre; the manured plots (No. 2) a lime requirement of 2,197 
pounds ; the leached manured plots, 3,085 pounds ; and the plots re- 
ceiving the leachings from the manure (No. 5), a lime requirement 
of 2,024 pounds per acre. 

The constituents and character of the materials incorporated in the 
soil in this test have produced different effects. The manure, robbed 
of its salts and of its soluble nitrogenous and other organic sub- 
stances, has produced the greatest acidity. The soluble part of the 
manure, including both the organic and inorganic constituents, has 
produced conditions which cause a less acid condition than the un- 
treated soil. The degree of acidity of the plots receiving the whole 
manure (No. 2) is in harmony with the condition of the plots receiv- 
ing the separate parts of the manure (Nos. 3 and 5), for its hme 
requirement is less than the plots receiving the leached manure (No. 
2) and more than the plots receiving the leachings (No. 5). This 
substantiates the idea that the straw and insoluble part of the manure 
by its decomposition in the soil has produced conditions which in- 
crease its acidity, and the soluble inorganic and organic portions have 



WALLKR: SKLF-l'OLLI nation IN MAIZK. 



35 



produced the reverse conditions, decreasinj^ the normal acidity of 
the soil. 

SUMMARY. 

In an experiment ^rowinc^ wheat, rye, clover, timothy, corn, cow- 
peas, and potatoes, conducted on a heavy silty clay loam at Arlington, 
Va., calcium sulfate, ferrous sulfate, manganese sulfate, potassium 
sulfate, and potassium sulfide added singly to the soil annually for 
five years increased its acidity. 

Magnesium carbonate decreased the acidity of the soil. Soil fer- 
tilized with sodium nitrate was less acid than the untreated soil or 
soil fertilized with acid phosphate or potassium sulfate. Acid phos- 
phate fertilization increased the acidity of the soil, but not as much 
so as potassium sulfate. 

Organic materials afifected the soil differently as to causing acidity. 
Starch caused increased acidity ; stable manure slightly increased 
acidity, which was still greater with manure leached of its soluble 
organic and inorganic substances. The leachings from manure pro- 
duced less acidity than the untreated soil and less than the whole 
manure or leached manure. The nature of the decomposition of the 
organic material in the soil and the character of the life processes in 
the soil affects the influence of such substances on soil acidity. 

Bureau of Plant Industry, 
U. S. Department of Agriculture, 
Washington, D. C. 



A METHOD FOR DETERMINING THE PERCENTAGE OF 
SELF-POLLINATION IN MAIZE.^ 

A. E. Waller, 
Ohio State University, Columbus, Ohio. 

In genetic studies or in field practice the seed-corn grower often 
finds that a small culture, if the number of individuals is not in- 
creased, soon consists of more or less closely related individuals 
and a decrease in yield and general vigor of the plant is the normal 
result. While it is generally recognized that there is some self- 
pollination in a field of corn, it would be interesting to know just how 
frequently it occurs naturally. So far as the writer knows, no method 

1 Received for publication June 26, 1916. 



36 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

has been devised for determining the amount of crossing or selling 
that takes place under field conditions. The fact that the number of 
pollen grains given off from a single plant is estimated in the millions 
suggests the great opportunity for crossing. 

The appearance of dwarfs, plants lacking in chlorophyll, and vari- 
ous other abnormalities supposed to be recessive and carried latent, 
being hidden by normal (dominant) conditions, can be explained on 
the Mendelian basis only upon the supposition of a previous self- 
pollination. This is because there is always a greater chance of 
bringing out recessive characters when closely related individuals are 
paired. 

In the experiment here described the method used to determine 
the amount of selling was quite simple. It consisted of planting hills 
of white corn in a field of yellow. At the time the corn tasseled two 
of the three plants in each hill of the white were detasseled. The 
ears were harvested from the plants upon which the tassels were 
allowed to remain. Those kernels which had been cross-pollinated 
were yellow, due to xenia, while those which were self-pollinated 
were of course white. 

In order to insure normal crossing between the white and yellow 
varieties it was necessary to procure two varieties which mature about 
the same time. For this purpose a selection of Reid's Yellow Dent 
developed at the Ohio State University and Wing's Hundred-Day 
White were used. So far as time of shedding pollen was concerned 
this was a fortunate combination. 

The hills containing the white corn were spaced every eleventh row 
east and west, a distance of 38^ feet from hill to hill, and every 
tenth row north and south, a distance of 35 feet from hill to hill. 
Naturally, climatic conditions would afifect the distance apart the 
hills could be spaced without danger of pollen infection from one 
white hill to another. When thoroughly dried, the numbers of yellow 
and of white kernels were counted. The results are shown in 
Table i. 

Only one pair of endosperm characters was contrasted in this 
experiment, i. e., white and yellow. Had two contrasting pairs such 
as color and composition of endosperm been employed the classifica- 
tion of the grains would have been much easier. The yellow was 
often so weak that it was difficult to distinguish between white and 
yellow kernels. This was perhaps because in the endosperm forma- 
tion there was only one set of determiners for yellow, that brought 
by the second nucleus of the pollen grain, to two sets of determiners 
for white, those which were contained in the double endosperm 



WALLHR : SKLF-l'Ol.l.l NA IION IN MAI/.i;, 



37 



nucleus. In endosperm formation in mai/x' the second nucleus from 
the pollen fuses with the two polar nuclei in tri])lc fusion; if the 
male nucleus is the bearer of determiners for a dominant character 
expressing^ itself in the endosperm, the phenomenon called xenia then 
results. In this experiment if, instead of using^ one recessive endo- 
sperm character, white color, a second such as suj^ary condition of 
the endosperm had been combined with it, the chance for making a 
mistake in the classification and counting would largely have been 
eliminated. However, it would have been difficult to obtain a variety 
of white sweet corn maturing at the same time the field corn matured, 
and a series of plantings would have been necessary. 



Table i. — Number of yellow {cross-pollinated) and of white (selfed) kernels 
on ears of white corn, with percentage of selfed kernels. 



Ear 


Number 


Number 


Percentage 


Ear 


Number 


Number 


Percentage 


No. 


of yellow 


of white 


of selfed 


No. 


of yellow 


of white 


of selfed 


kernels. 


kernels. 


kernels. 


kernels. 


kernels. 


kernels. 


I 


All 








20 


651 


50 


7. II 


2 


379 


112 


23.10 


21 


691 


23 


3.20 


3 


704 


5 


0.70 


22 


517 


3 


0.58 


4 


214 


7 


3-16 


23 


428 


14 


3.16 


5 


564 


15 


2.58 


24 


372 


82 


18.06 


6 


369 


13 


3.40 


25 


653 


28 


4. II 


7 


All 








27 


671 


3 


0.44 


8 


633 


29 


4-38 


28 


668 


15 


2,19 


9 


496 


21 


4.06 


29 


793 


8 


1. 00 


10 


All 








30 


383 








II 


629 








31 


806 


8 


0.98 


12 


678 


26 


3-69 


32 


664 


55 


7.64 


13 


601 


29 


4.60 


33 


744 


24 


3.12 


14 


496 


143 


22:37 




773 


78 


9.16 


15 


66s 


21 


3.06 


I' 
35 


614 


31 


4.80 


i6 


425 


63 


12.93 


36 


529 


26 


4.68 


17 


695 


II 


1-55 


37 


616 


34 


5-23 


i8 


586 


89 


13-18 


38 


344 


54 


13-50 


19 


639 


21 


3.18 


























Averae^e 




5-13 

















Conclusions. 

1. The average amount of self-pollination obtained under these 
conditions was 5.13 percent, but this figure may not represent the 
percentage of self-pollination taking place under all field conditions, 
since humidity and wind are factors which determine the zone of 
infection between hills. To obtain significant figures the experiment 
should be repeated for several years. 

2. Better examples of the effect of xenia can be obtained when, 
more than one contrasting pair of characters are employed. 

Ohio State University, 
Columbus, Ohio. 



38 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



A METHOD FOR DETERMINING THE VOLUME WEIGHT OF 
SOILS IN FIELD CONDITION.^ 

Charles F. Shaw. 

In physical and chemical soil studies the determinations of soil 
moisture, lime, alkali, plant-food ingredients, etc., are most commonly 
and conveniently reported as percentages by weight, based on water- 
free soil. When it is desirable to know the total amount of material 
in the soil studies, it becomes necessary to know the weight of a unit 
volume of the soil or to know the volume weight. In irrigated farm- 
ing this knowledge is especially valuable. The wilting point and the 
total water-holding capacity of the soil can be determined, and from 
these the avaliable water can be calculated. But as irrigation water 
is purchased by volume measurements and is expensive, we need to 
know the volume of water in the soil and the added volume needed 
to bring the moisture to the desired amount. 

There are several methods of determining the volume weight, or 
apparent specific gravity, of soils. These may be classed under two 
general heads, those where the measurements are made on disturbed 
soil and those where the measurements are made on soils in their 
natural field condition. The first class is best exemplified by the 
common laboratory method in which the air-dry or oven-dry soil is 
placed in a brass cylinder or other suitable container, compacted by 
some arbitrary method, and the weight and volume determined. Of 
the second class the most common is the so-called King tube method, 
in which a tube with a suitable cutting edge is forced into the soil for 
a measured distance and the weight of the enclosed soil used as the 
basis for determining the volume weight. 

The first method has the obvious weakness of dealing with a dis- 
turbed soil, and through manipulation of the soil, the results obtained 
may show a volume weight much greater or much less than that 
which existed in the original undisturbed or field condition. The 
second method gives a closer approximation to the true condition but, 
particularly in heavy soils, the pressure needed to force the tube into 
the soil causes appreciable compaction of the soil ahead of the tube, 

1 Presented at the ninth annual meeting of the American Society of Agron- 
omy, Washington, D. C, November 13, 1916. Contribution from the Labo- 
ratory of Soil Technology, University of California. 



s 1 1 A vv : v( ) L u M !•: w I d 1 1 r ov so i ls . 



39 



modifying the results. In stony or gravelly soils it is almost im- 
possible to use the tubes with any degree of satisfaction. 

The method about to be described is not new. It was worked out 
by the author in 1908 at the Pennsylvania State College, and a brief 
description was published in 1909 by B. E. Brown and W. F. Cree.^ 
A somewhat similar method is mentioned by Stevenson,^ and one 
almost identical was worked up by Dr. Rudolf Trnka.* This 
method has been termed the paraffine immersion method. While it 
is difificult to handle, it offers a means of making very accurate de- 
terminations on all but the more sandy soils. In our experiments 
blocks of soil about i foot square and 6 inches deep were carefully 
cut from the soil mass and lifted on broad-bladed spatulas to a table, 
where the block was divided into smaller blocks, from 4 to 6 inches 
square and 6 inches deep. In subdividing the block great care was 
taken to preserve the full 6-inch depth and the general cross section, 
but the exact shape or size of the cross section was not so important. 
From the base soil secured in preparing these blocks, small samples 
were taken for moisture determinations, in order to correct the 
weight to the water- free basis. 

The blocks were placed on a weighed support or dipping device and 
carefully .weighed, then dipped several times in melted paraffine. The 
coating of paraffine should be heavy enough to stand some necessary 
handling and should be waterproof. After cooling, the blocks were 
weighed to determine the amount of paraffine taken up. The paraf- 
fined soil was then weighed in water, the difference in weight show- 
ing the weight of the water displaced. In some of our work, in- 
stead of weighing in water, the volume was determined by actual 
measurement of water displacement, either by placing the block in a 
container of known capacity and measuring the amount of water 
needed to fill, or by immersing in a vessel full of water and measur- 
ing the water that flowed from the outlet. 

After the weighing or measuring was completed the block was 
broken up to recover the paraffine and to note any unusual features, 
such as worm holes, excessive quantities of roots or stones, or any 
other feature that would make the sample exceptional. In some 

2 Brown, B. E., Maclntire, W. H., and Cree, W. F. Comparative physical and 
chemical studies of five plats, treated differently for twenty-eight years. In 
Ann. Rpt Pa. Agr. Expt. Sta. 1909-10, pp. 96, 97. 1910. 

3 Stevenson, W. H. A new soil sampler. Iowa Agr, Exp. Sta. Bui. 94, 31 p., 
9 fig. 1908. 

* Trnka, Rudolf. Eine Studie iiber einige physikalischen Eigenschaften des 
Bodens. Internat. Mitt, fiir Bodenkunde,. bd. 4, heft 4/5, p. 363-387. 1914. 



40 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



cases, where an unusual number of pebbles was present, they were 
collected, washed, dried, and weighed. Their volume was then de- 
termined by immersing in a graduated cylinder and suitable deduc- 
tions made from the soil figures. 

From the data thus obtained the volume weight was calculated by 
the following formula: 

V — {N — M) — R' — V olume weight, where 

M = weight of soil in field condition ; 
N = weight of soil and paraffine ; 
V = volume of soil and paraffine ; 
P = percentage of water in the soil ; 
R — weight of stone in the soil; 
R' = volume of stone in the soil ; and 
.9 = specific gravity of the paraffine. 
The volume and weight of the dipping device being constant, they are not 
included in M, N, and V. 

The results shown in Table i were obtained at the Pennsylvania 
State College in 1908 and 1909 by using this method. 



Table i. — Weight per cubic foot of soil as determined by the paraffine immer- 
sion method. 



Plot No. 


Treatment. 


No. of trials. 


Volume weight. 


Weight per cubic 
foot, lbs. 


I 


Check 


6 


1.3366 


83.479 


2 


Dried blood 


3 


I.341O 


82.910 


3 


Dissolved bone black 


2 


1.2850 


80.250 


4 


Muriate of potash 


3 


1.2500 


78.160 


S 


N - P 


4 


1.2000 


75-005 


6 


N-K 


4 


I.3IOO 


82.100 


9 


N-P-K 


8 


1.367s 


85.310 


15 


P-K 


3 


I.4IOO 


88.180 


Average 




33 


I.3128 


81.923 



In making calculations for fertilizer treatments on these plots, the 
experiment station had been using 70 pounds as the weight per cubic 
foot of this soil. 

In the summer of 191 6, in cooperation with Mr. O. W. Israelson, 
of the Division of Irrigation Investigations of the California station, 
measurements were made on the Willows clay, a compact silty clay 
having a dense structure. Water percolates into this soil very slowly, 
making irrigation very difficult. In this work extreme care was 
taken to get accurate results, although we were somewhat handi- 



siiAw: voLUMi: \vi:i(;iiT ov soii.s. 



41 



capped by the necessity of doing the work in the field, over 100 miles 
from the laboratory. The samples were taken at about 6-inch depths, 
thouj^h the necessity of leveling off the column before taking the next 
sample reduced the depths somewhat. Samples were taken to a 
depth of 5 feet, being cut from the side of a large hole which was 
dug to facilitate getting at the soil. The results are shown in 
Table 2. 



Table 2. — Volume weight of Willows clay at various depths, as determined by 
the laboratory and the paraffine immersion methods. 







Volun 


le weight. 


Depth of soil. 


inches. 


Laboratory method (average of 
12 determinations). 


Paraffine immersion method (average 
of 3 determinations). 


2-71 






1. 671 d=.004 


8-13 






1.702 ±.008 


15-20 




1.380 ±.007 


i.784±.oo7 


22-27 






1.802 ±.002 


28-33 . 






1.802 ±.007 


35-40 V 






1.792 ±.000 


42-47 
48-53 




1.320 ±.004 


i.757±-oi2 

1. 741 ±.010 


54-60 






1. 541 ±.028 




1.350 ±.008 


i.733±-035 



It will be noted that the results in this case agree with those from 
Pennsylvania in indicating that the soil in its natural field condition 
is much heavier than is indicated by the laboratory method. In the 
article already mentioned, Trnka also cites results showing the natural 
soil condition to be consistently more dense than the laboratory- 
methods show. There seems to be sufficient evidence that the older 
methods gave results that were quite different from those existing in 
the field, emphasizing the necessity for more careful study by more 
accurate methods. 

During 191 5 and 1916 Mr. O. W. Israelson and others in the 
Division of Irrigation Investigations at the California station, work- 
ing in cooperation with the Farm Irrigation Investigations of the U. 
S. Office of Public Roads and Rural Engineering, developed a new 
method for rapid approximate determinations of volume weight of 
soils in the field condition. This will be more fully discussed in a 
later publication, and will be but briefly mentioned here. 

The apparatus consists of a straight-sided soil auger of the post- 
hole type — a cylinder with two side cutting edges, — a long tube of thin 
rubber of slightly smaller diameter than the auger, a large graduated 
cylinder, and a hand pump. A hole is bored to the desired depth, 



42 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

all the soil being carefully collected, sampled for moisture, and 
weighed. The rubber tube is carefully inserted in the hole, and 
water is measured in until the tube is filled to the top of the hole. 
The water is then pumped out, and the tube removed and dried. The 
process is repeated until the final desired depth is reached. This 
method, when checked against the paraffine immersion method, gave 
very close results. On the Willows soil the rubber tube method 
showed a volume weight of 1.744 ± .010 while the paraffine immer- 
sion method gave 1.733 ± -^35- Checked against the laboratory 
method the rubber tube method gave the results shown in Table 3. 



Table 3. — Volume weights of various soils as determined by the laboratory and 
the rubber tube methods. 



Soil. 


Volume 


weight. 


Laboratory method. 


Rubber tube method. 


Fine sandy loam 


1.20 


1.280 


Loam 


1.22 


1. 210 


Clay 


1. 13 


I.2S7 


Clay loam 


I.I8 


1.398 


Clay 


1-34 


1.642 


Silty clay 


1.38 


1. 741 


do. 


1.35 


I-7S0 


Fine sandy loam 


1.28 


1. 112 


Loam 


1.20 


1.289 


Clay loam 


1-35 


1.272 



These figures show that there usually is a marked difference be- 
tween the results obtained by the two methods, and further empha- 
size the importance of a more careful study of the problem. 
Laboratory of Soil Technology, 
University of California, 
Berkeley, Cal. 



AGRONOMH AI'l'AlRS. 



43 



AGRONOMIC AFFAIRS. 

THE SOCIETY IN 1917. 

The American Society of Agronomy is now entering on its tenth 
year. Organized on December 31, 1907, it has grown to a society 
of more than 600 members in nine years. Its membership has in- 
creased most rapidly during recent years, largely because the issu- 
ance of a journal has greatly increased its interest to agronomic 
workers. With the greater frequency of issue incident to the cur- 
rent volume, this interest should be heightened and the membership 
still largely increased. The Society should have every agronomic 
worker in the United States and Canada on its roll, with many repre- 
sentatives in other countries as well. By the end of its tenth year it 
should have at least 750 members. If only one in four of the pres- 
ent members brings in a new member during the year the 750 mark 
will be more than passed. The Secretary is doing what he can to 
bring the Society to the attention of nonmembers, by means of cir- 
cular letters and sample copies of the Journal, but a personal invi- 
tation is a far more effective means of inducing men to join. 

With the increasing membership, each member receives more from 
the Society. In 191 6, 25 percent more pages of agronomic matter 
were printed in the Journal than in the previous volume. There 
will be a further increase in 1917. Of more importance, however, 
is the greater frequency of publication. At a meeting of the Execu- 
tive Committee in November, the publication of nine numbers in 
191 7 was authorized, these to appear monthly from January to 
May and from September to December. It is the plan to have the 
Journal appear about the fifteenth of each month. This more fre- 
quent and regular publication should add to its value and usefulness. 
Papers on agronomic subjects are solicited, especially those from 2 
to 8 pages in length. Longer papers are frequently accepted and 
are always useful, but it is the aim of the Editor to have matter in 
the Journal presented in crisp, concise form. A" short paper will 
often be read in its entirety, while a longer one will be read by title 
only and then put aside for reading at a more convenient time, with 
the result that it is not read at all. 

The annual meeting of the Society in Washington in November 
was a markedly constructive one. The committees of the Society 



44 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

are doing excellent work and are formulating plans of much value. 
Agronomic workers everywhere should know of this work and 
should help in it. This knowledge and cooperation can be gained only 
by membership in the Society. The series of papers on agronomic 
terminology contributed by the committee on that subject will be con- 
tinued from time to time during the year, and other series of in- 
terest and value are projected for the Journal. If we work to- 
gether, we can make 191 7 by far the best year that the American 
Society of Agronomy has ever known. Will you not ask at least 
one man to join the Society today? 

ANNUAL DUES OF MEMBERS. 

Two changes were made in the by-laws of the Society at the No- 
vember, 1916, meeting, both of which affect the payment of annual 
dues. These are recorded in the minutes of the meeting published 
in the November-December Journal, but for convenience are re- 
produced here, the changes being printed in italics: 

1. The annual dues for each active and associate member shall be $2.00, and 
for each local member 50 cents, which shall be due and payable on January i 
of the year for which membership is held. 

2. The Journal of the American Society of Agronomy shall not be sent to 
any member whose dues are not paid by April i of the year for which member- 
ship is held. 

According to these by-laws, dues are now payable on January i 
instead of April i, and membership in the Society lapses unless 
dues are paid within the calendar year for which membership is 
held. In the past, lapsed members received the Journal through 
an entire year and until April i of the year following, for which the 
Society received no payment. Now, the delivery of the Journal 
is stopped on April i of the year for which membership is held, if 
dues are not paid by that time. This seems to be an eminently fair 
arrangement, as those who are not sufficiently interested in the 
Society to pay its annual dues certainly have little interest in the 
Journal, while the expense of sending the magazine to them is not 
a justifiable charge against those who pay the membership fee regu- 
larly. Members are urged, therefore, to remit their dues for 191 7 
promptly to the Treasurer, Prof. George Roberts, Experiment Sta- 
tion, Lexington, Ky. Prompt remittance will not only lessen the 
work of that faithful officer but will insure prompt and regular de- 
livery of the Journal through 191 7. 



AGRONOMIC AFFAIRS. 



45 



MEMBERSHIP CHANGES. 

The membership of the Society at the end of 1916 was 586. 
Since that time 24 new members have l)een added, i has been rein- 
stated, and 2 have resigned, making a net gain of 23. The member- 
ship of the Society, therefore, is now 609. The names and addresses 
of these men, with such changes of address as have been reported to 
the Secretary, follow. The address of one member, Robert K. 
Bonnett, wdiich was incorrectly given in the address list published in 
the preceding number of the Journal, is also printed here. 

New Members. 

Albert, A. R., Dept. of Soils, College of Agriculture, Madison, Wis. 

Brewer, Herbert C, The Barrett Company, 17 Battery Place, New York, N. Y. 

Briggs, Glen, 318 West St., Stillwater, Okla. 

Brockson. W. I., Agr. Expt. Sta., University of Illinois, Urbana, 111. 
Curtis, Harry P., 1103 West Springfield Ave., Urbana, 111. 
Douglas, J. P., 402 East Chalmers St., Urbana, 111. 

Gray, Samuel D., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 
Hanson, Lewis P., Dept. of Soils, College of Agriculture, Madison, Wis. 
Harlan, Harry V., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 
Haskell,, E. S., Farm Management, U. S. Dept. Agr., Washington, D. C. 
Hodgson, E. R., Agr. Expt. Sta., Blacksburg, Va. 

Hulbert, Harold W., Farm Crops Dept., Iowa State College, Ames, Iowa. 

Huston, H. A., German Kali Works, 42 Broadway, New York, N. Y. 

Joslyn, H. L., Craven Co. Farm Life School, Vanceboro, N. C. 

Kelly, E. O. G., Wellington, Kans. 

Kemp, Arnold R., 402 Chalmers St., Champaign, 111. 

Longman, O. S., Olds Agricultural School, Olds, Alberta, Canada. 

MoRisoN, A. T., Dept. Agronomy, Agr. Expt. Sta., Urbana, 111. 

Moynan, John C, Cereal Husb. Dept., Macdonald College, Quebec, Canada. 

Northrop, Robert S., Box 442, Redding, Cal. 

Osenbrug, Albert, Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 
Park, J. B., Farm Crops Dept., Ohio State University, Columbus, Ohio. 
Purington, James A., Mass. Agricultural College, Amherst, Mass. 
WiLLARD, C. J., 919 Nevada St., Urbana, 111. 

Member Reinstated. 
Marbut, C. F., Bur. Soils, U. S.. Dept. Agr., Washington, D. C. 

Members Resigned. 
Carnes, Homer M., North Powder, Oreg. 
Sweet, Carl, Dominion Seed Branch, Regina, Sask., Canada. 

Changes of Address. 
Atwateh, C. G., Agr. Dept. The Barrett Co., 17 Battery Place. New York, N, Y. 
Bonnett, Robert K., Kans. State Agr. College, Manhattan, Kans. 
Woodard, John, 381 Paisley Road, Guelph, Ont., Canada. 



46 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



NOTES AND NEWS. 

W. A. Albrecht has been elected instructor in soils at the Univer- 
sity of Missouri. 

William W. Baer has been appointed assistant chemist at the New- 
York state station for work in the agronomy department. 

E. C. Bowers has been elected instructor in agronomy at the Vir- 
ginia Polytechnic Institute. 

N. I. Butt is now assistant in agronomy at the Utah college and 
station. 

H. J. Conlin has been appointed assistant in soils at the Ohio 
station. 

W. B. Ellett, chemist of the Virginia station, has been appointed 
professor of agricultural chemistry in the Virginia college in addi- 
tion to his station work. 

R. A. Kinnaird, who for the past year has been teaching agricul- 
ture at the normal school at Maryville, Mo., is now extension in- 
structor in soils in the University of Missouri. 

J. D. Luckett, formerly of the Purdue University station, is now 
on the editorial staff of the Experiment Station Record, where he 
is giving special attention to publications on field crops. 

A. M. Peter, for the past several years chemist of the Kentucky 
station, has recently been elected director of that station. 

M. N. Pope, instructor in biology and agriculture in the St. Paul 
central high school last year, is now teaching the same subjects in the 
State Normal School at Eau Claire, Wis. 

Phil E. Richards, assistant in agronomy and graduate student at 
Ohio State University during the past year, is assistant in soil re- 
search at the Maryland station. 

Newell S. Robb, formerly assistant agronomist in the University 
of Idaho, since June 15 has been county agent in Lane County, 
Oregon, with headquarters at Eugene. 

Nickolas Schmitz, agronomist of the Maryland station, has been 
elected agronomist in the extension division at the Pennsylvania 
State College. He will have charge of all extension work in agron- 
omy in Pennsylvania. 



AGRONOMIC AFFAIRS. 



47 



Clinton D. Smith, former director of the Arkansas, Minnesota, 
and Michigan stations and for five years president of the Louis 
Queiros school of agriculture at Sao Paulo, Brazil, died at Buffalo, 
N. Y., August 5, at the age of 62 years. 

W. H. Smith, state superintendent of education, has been appointed 
president of the Mississippi Agricultural College, succeeding George 
R. Hightower on September 15. 

George Stewart, assistant in soils at the Utah college and station, 
is pursuing graduate study at Cornell University. 

F. W. Stemple, formerly instructor in agronomy at Ohio State 
University, is now professor of agronomy and agronomist of the 
West Virginia college and station. 

R. W. Thatcher has recently been appointed assistant director of 
the Minnesota station, in addition to his other duties. Dr. Thatcher 
is also acting president of the newly organized American Associa- 
tion for the Promotion of Technical Training in India, a society 
which is planned to be of assistance to Hindoo students in the 
United States and to aid them in developing industrial education on 
their return to India. 

John B. Wentz, who has been engaged in post-graduate work at 
Cornell University during the past year, is now associate professor 
of farm crops at the Maryland State College. 

Leroy D. Willey, formerly assistant in dry-land agriculture on the 
Cheyenne Experiment Farm, Archer, Wyo., is now superintendent 
of the newly established Sheridan Field Station at Sheridan, Wyo. 

The Agricultural Digest, the first issue of which appeared early 
in the summer, is the official publication of the National Agricultural 
Society. The Digest will contain reviews of the leading articles in 
the various farm papers and of federal and state bulletins on agri- 
culture, and also original articles on timely topics. The National 
Agricultural Society, " a national organization for the promotion of 
agriculture," was organized on April 27, 1916. Hon. James Wilson, 
former Secretary of Agriculture, is president; G. Howard Davidson, 
chairman of the executive committee ; P. C. Long, secretary ; and 
Walter A. Johnson, treasurer. The headquarters of the Society are 
at 2 West 45th Street, New York. If financial strength is a prime 
requisite, the society should be an immediate success, for its board of 
directors contains such names as those of Theodore N. Vail, T. Cole- 
man du Pont, WiUiam H. Moore, and Samuel Insull. While the 



48 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

interest of these men in agriculture is unquestioned, their fame is due 
much more largely to their commercial successes. 

Appropriations for the U. S. Department of Agriculture. 

The appropriations act for the Federal Department of Agriculture 
for the fiscal year ending June 30, 191 7, which was approved by the 
President on August 11, set aside nearly $25,000,000 for the year's 
work. This is an increase of almost $2,000,000 over the appropria- 
tion for the previous year. The new legislation carried by the bill 
includes the U. S. grain standards act and the U. S. warehouse act. 

The increases in appropriations for the more important bureaus 
and offices were divided as follows: Weather Bureau, $81,000, prin- 
cipally for increased weather and storm- warning service in Panama 
and Alaska; Bureau of Animal Industry, $435,000, a large part of 
which is for control of contagious diseases of animals and the 
eradication of the Texas fever tick ; Bureau of Plant Industry, $398,- 
which is for control of contagious diseases of animals and the 
balance is distributed over numerous projects; Forest Service, $996,- 
000, for the purchase of eastern lands for forest purposes and for 
road construction in national forests; Bureau of Soils, $175,000, for 
investigating methods of obtaining potash on a commercial scale; 
Bureau of Biological Survey, $132,000, for the destruction of preda- 
tory animals ; States Relations Service, $148,000, principally for 
demonstration work ; and Office of Markets and Rural Organiza- 
tion, $388,000, principally for the establishment of a market news 
service on live stock, meats, fruits, and vegetables. The principal 
decrease in the bill is the cutting in half of the $2,500,000 emergency 
appropriation available last year for the eradication of the foot and 
mouth disease. This emergency appropriation is available only as 
necessary to combat outbreaks of foot and mouth or other con- 
tagious animal diseases. 

Other acts of the Congress which adjourned on September 9 in- 
clude the following increases in appropriations for agricultural pur- 
poses over those of the previous year : Smith-Lever extension fund, 
$500,000; printing fund, $100,000; cooperative construction of rural 
post roads, $5,000,000; and cooperative construction of roads and 
trails in the nationals forests, $1,000,000. 



JOURNAL 

OF THE 

American Society of Agronomy 



Vol. 9. February, 1917. No. 2. 



THE SOIL MULCH.i 

L. E. Call and M. C. Sewell. 

Introduction. 

It is a matter of common belief and teaching that the soil is culti- 
vated principally to maintain an earth mulch in order to prevent 
evaporation and to conserve moisture. The purpose of this paper is 
to review the experiments in soil mulches which have been conducted 
and to present certain results which do not conform to accepted 
teachings. 

It has usually been taught (i) that the soil moisture is capable of 
mo\ ement through capillary force ; (2) that this force is a result of the 
forces of cohesion and surface tension, when the latter is not in a 
state of equiHbrium;^ (3) that the phenomena of surface tension are 
due to the existence of molecular forces ; and (4) that moisture within 
the soil exists in the form of films about the soil particles. 

The phenomena are illustrated by the suspended drop of water. 
The particles in the interior of the drop are attracted equally in all 
directions by the other particles of the liquid. This is cohesion or the 
attraction of Hke molecules of matter. A molecule on the surface of 
the drop is not attracted equally on all sides, since the molecules of 
gas surrounding the drop have less attraction for the particles than is 
exerted by the particles within the liquid. Hence the resultant at- 
traction is inward, and there is formed a membrane, as it were, of 

^ Contribution from the Kansas Agricultural Experiment Station. Presented 
by the junior author at the ninth annual meeting of the American Society of 
Agronomy, Washington, D. C, November 13, 1916. 

2 Briggs, Lyman J. The mechanics of soil moisture. U. S. Dept. Agr., Bur. 
Soils Bui. 10. 1897. 

49 



50 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

uniform tension. Thus we have the soil moisture, except in cases of 
saturated soil, in the form of films. These films of water about 
respective soil particles grouped together exert a pressure dependent 
upon the capillary angle, which is the angle formed between the liquid- 
solid and the liquid-gas surface. The angle is dependent upon the 
form of the film. Considering three particles of soil enclosed with 
films in contact, and one of these films connecting the water held in 
two adjacent capillary spaces containing different quantities of water, 
the capillary angle between the film with the less water and the joining 
film will have the greater curvature and pressure outward. Hence 
the angle being greater and outward pressure being greater, water will 
be drawn over until equilibrium exists between the pressure exerted 
by the water in the two capillary spaces. Such movement exists be- 
tween all of the soil particles, the tendency being to bring about a 
condition of equilibrium between the pressures of adjacent films. If 
the surface tension of water in the upper layers of soil is increased, 
water is drawn toward that point. This in brief constitutes the 
theory of the movement of soil moisture. 

The Theory of Soil Mulches. 
The theory of the soil mulch in preventing or checking evaporation 
of soil moisture is that by loosening the surface by cultivation the soil 
particles are removed from intimate contact with one another, the 
films lose most of their water by evaporation, and the movement of 
water to the surface is thus prevented. In other words, the dry layer 
of soil acts as a blanket in checking evaporation. 

Review of Literature. 
Laboratory experiments in which columns of soil in glass or brass 
containers are kept in contact with free water,^ and tank or lysimeter 
experiments in which a block or column of soil is kept in contact with 
a water table have fully demonstrated capillary movement of moisture 
for such conditions.* But capillary movement without the presence 

3 King, F. H. Textbook of the physics of agriculture, p. 161-170, 186. 1910. 

Hilgard, E. W. Soils ; their formation, properties, compositions, and rela- 
tions to climate and plant growth in the humid and arid regions, 1st ed., p. 203. 
The Macmillan Co., New York. 1902. 

* Fortier, Samuel. Soil mulches for checking evaporation. In U. S. Dept. 
Agr. Yearbook 1908, p. 465-472. 1909. 

King, F. H. Investigations relating to soil moisture, hi Wis. Agr. Expt. 
Sta. 8th Ann. Rpt'., p. 100-134. (1891) 1892. 

Investigations of soil management. U. S. Dept. Agr., Bur. Soils Bui. 

26, p. 198-205. 1905. 

Willard, R. E., and Humbert, E. P. Soil moisture. New Mex. Agr. Expt. 
Sta. Bui. 86. 1913. 



CALL SEWKLL: Till-: SOIL MULCH. 



51 



of a water tabic (with a dry subsoil) has not been demonstrated. In 
the Great Plains region the latter eondition is the more common, since 
on the uplands a water table is seldom encountered at depths of less 
than 60 feet. Because of the winds and dry atmosphere, the rate of 
evaporation also is higher than in more humid climates. Since the 
rate of evaporation may be more rapid than the rate of capillary move- 
ment to the surface, a natural air-dry mulch may be formed without 
cultivation. Hence, in such conditions, cultivation to conserve mois- 
ture might be of little value. 

Burr^ found at the North Platte (Nebr.) station that no water was 
drawn from the fifth foot to replace that which had been used by 
alfalfa on bench land underlaid with sheet water from 17 to 21 feet 
below the surface. In a period of seventeen months the moisture 
content of the fourth and fifth feet remained about constant. Burr 
concluded that "the plant roots to obtain water, extend themselves 
into the soil zone where available water is present, rather than depend 
upon the water being brought to them by capillarity." 

This conclusion is supported by the work of Miller^ in his study of 
the root systems of corn, kafir, and milo at the Garden City (Kans.) 
Branch Experiment Station during the seasons of 1914 and 191 5. It 
was found that the root systems of the crops by the latter part of 
A_ugust had extended themselves throughout the sixth foot of soil. 
From moisture determinations which were made a few days before 
or after the isolation of the various root systems, it was concluded 
that " the results of these experiments for both seasons seem to show 
that there was little if any depletion of the soil moisture below the 
depth to which the roots penetrated." 

Alway^ studied the growth of plants and their use of moisture when 
grown in water-tight cylinders at the Nebraska station. He found 
that the moisture becomes available to the plants by the development 
of roots to moist soil, and that but little moisture is elevated to the 
roots by capillarity. 

The capillary movement of moisture was studied at the New Mexico 
station^ by lysimeter experiments. The maximum rise of moisture 
from water through sandy loam was found to be 32 inches and 

^ Burr, W. W. The storage and use of soil moisture. Nebr, Agr. Expt. Sta. 
Research Bui. 5. 1914. 

^ Miller, Edwin C. Comparative study of the root systems and leaf areas 
of corn and the sorghums. In U. S. Dept. Agr., Jour. Agr. Research., v. 6, 
no. 9, p. 311-331- 1916. 

7 Alway, F. J. Studies on the relation of the nonavailable water of the soil 
to the hygroscopic coefficient. Nebr. Agr. Expt. Sta. Research Bui. 3. 1913. 

« Willard, R. E., and Humbert, E. P. Loc. cit. 



52 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY, 

through adobe clay 50.5 inches. The movement of moisture from 
wet to dry soil was very slight, although moisture rose 30 inches from 
wet soil to the roots of a crop of wheat. 

Barker,^ of the Nebraska station, reported that the loss of water 
because of direct evaporation from the surface of the soil is very 
small after the water becomes thoroughly distributed. Young,^^ of 
the same station, found that the soil mulch is not more effective than 
an unmulched soil in retarding the evaporation of the moisture that 
is well established in the soil; that if a hard layer of soil dries out to 
the depth of 2 or 3 inches it will act as a mulch ; and that the loss of 
water from the soil is largely due to transpiration from plants. This 
last conclusion is supported by Alway,^^ who found in studies con- 
ducted at the Nebraska station that the loss of water from the subsoil 
of dry lands under crop seems to take place almost entirely through 
transpiration. In the absence of plants, the loss from the subsoil is 
small. 

A review of tillage experiments^^ shows that no field experiments 

* Barker, P. B. The moisture content of field soils under different treat- 
ments. In Nebr. Agr. Expt. Sta. 25th Ann. Rpt'., p. 106-110. (1911) 1912. 

10 Young, H. J. Soil mulch. In Nebr. Agr. Expt. Sta. 25th Ann. Rpt, p. 
124-128. (1911) 1912. 

11 Alway, F. J. Loc. cit. 

12 Atkinson, Alfred, Buckman, H. O., and Gieseker, L. F. Dry farm mois- 
.ture studies. Mont. Agr. Expt. Sta. Bui. 87. 191 1. 

Barker, P. B. Loc. cit. 

Buckman, H. O. Moisture and nitrate relations in dry-land agriculture. 
Proc. Amer. Soc. Agron., 2: 131. 1910. 

Cardon, P. V. Tillage and rotation experiments at Nephi, Utah. U. S. 
Dept. Agr. Bui. 157. 1915. 

Jardine, W. M. Arid farming investigations. Utah Agr. Expt. Sta. Bui. 

TLQO. 1906. 

Linfield, F. B., and Atkinson, Alfred. Dry farming in Montana. Mont. Agr. 
Expf. Sta. Bui. 63. 1907. 

Lipman, C. B. Plowing and cultivating soils in California, Cal. Agr. Expt. 
Sta. Circ. 98. 1913. 

Loughridge, R. H. Moisture in California soils during the dry season of 
1898. In Cal. Agr. Expt. Sta. Ann. Rpt. 1897-8, p. 65-96. 1900. 

Schollander, E. G. Moisture conservation, /w Williston (N, Dak.) Subexpt. 
Sta. 3d Ann. Rpt, p. 61-63. IQIO- 

Shutt, Frank T. The moisture content of packed and unpacked soils. In 
Rpt. of the Chemist, Expt Farms Rpts. (Canada), p. 172. 191 1. 

Widtsoe, John A. The storage of winter precipitation in soils. Utah Agr. 
Expt Sta. Bui. 104. 1908. 

Irrigation investigations : factors influencing evaporation and transpi- 
ration. Utah Agr. Expt. Sta. Bui. 105. 1909. 

Young, H. J. Loc. cit 



CALL \' skwell: the soil mulc h. 



53 



on the same soil types have been conducted in wliich the moisture 
content of cultivated surfaces and surfaces bare of all vegetation but 
uncultivated have been compared, except the experiment of H. J. 
Young at the Nebraska station. Experiments conducted at the 
Kansas station at Manhattan and at the Garden City substation were 
planned to show this comparison. The writers believe that the data 
secured are of value in the consideration of economic tillage. 

Experimental Data on Soil Moisture. 

Experiments in which uncropped areas were (i) cultivated and 
(2) left uncultivated but kept free of weeds were begun at Man- 
hattan in 1909 and at Garden City in 1912. In both cases moisture 
determinations were made to a depth of several feet. The experi- 
ments at Manhattan were discontinued at the close of the 1909 season 
but were resumed again in 1914. 

EXPERIMENTS AT MANHATTAN. 

Soil samples were taken to a depth of 6 feet at Manhattan in 1909. 
The moisture determinations expressed in percentages and in inches 
of water are presented in Table i. The soil was a Marshall silt loam 
and the plots one twentieth acre in size. The uncultivated plot was 
kept bare of vegetation by scraping with a hoe and the cultivated plot 
was worked with a i -horse cultivator. 



Table i. — Moisture content (calculated to dry soil) of soil at various depths 
from cultivated and uncultivated plots at Manhattan, Kans., in May, 
July, and August, igog. 





May. 


July. 


August. 


Depth (feet). 


Cultivated. 


Uncultivated. 


Cultivated. 


Uncultivated. 


Cultivated. 


Uncultivated. 


Per- 
cent. 


Inches. 


Per- 
cent. 


Inches. 


Per- 
cent. 


Inches. 


Per- 
cent. 


Inches. 


Per- 
cent. 


Inches. 


Per- 
cent. 


Inches 


I 


28.42 


4.26 


27.89 


4.18 


26.07 


3-91 


24.33 


3.65 


23.06 


3-46 


20.55 


3-08 


2 


25.42 


4.14 


25-51 


4.16 


25.07 


4.09 


26.66 


4.34 


24.52 


3.99 


25.64 


4.18 


3 


21.54 


3.86 


21.87 


3-92 


21.40 


3.83 


22.25 


3.90 


21.04 


3.77 


20.61 


3.69 


4 


20.25 


3-81 


20.48 


3-86 


21.28 


4.01 


20.21 


3.80 


20.41 


3.84 


20.05 


3-77 


5 


20.02 


3.51 


20.82 


3.66 


20.53 


3.61 


29.60 


5.20 


20.29 


3.57 


19.54 


3-43 


6 


20.73 


3.48 


21.00 


3-52 


20.89 


3-67 


20.29 


3-40 


20.16 


3.38 


19.31 


3.24 


Total 




23.06 




23.30 




23.12 




24.29 




22.01 




21.39 

















During the season the mulched plots lost a total of 1.05 inches of 
water and the unmulched, 1.91 inches, the difference probably being 
within the experimental error. 

The experiment was resumed in a modified form in the spring of 



54 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

1914. Four plots were included, (i) cultivated 3 inches deep; (2) 
cultivated 6 inches deep; (3) not cultivated, weeds allowed to grow; 
and (4) not cultivated, but weeds removed by scraping. The avail-, 
able plant food as indicated by nitrate determinations in mulched and 
nonmulched plots was also determined. 

The size of the plots was 12 feet by 25 feet. The soil is known as 
a Marshall silt loam. The moisture equivalent of this soil is 27.52 



Table 2. — Moisture content (percentage of dry weight of soil) of mulched and 
unmulched plots at Manhattan, Kans., on various dates in 1914. 



Date of sampling. 


Treatment. 


Weeds. 


3-inch mulch. 


6-inch mulch. 


Bare surface. 


April 13 


22.36 


23.01 


19-34 


21.54 




18.30 


19.52 


17-93 


16.47 


June 4 


16.85 


17-83 


14.98 


19.38 


July 9 


14.42 


22.37 


22.78 


22.08 


July 23 


15-38 


18.22 


18.33 


19.62 


August 5 


13.88 


18.67 


17-45 


17.00 


August 25 


13.12 


16.87 


16.37 


15.13 


November 2 


15.08 


19.70 


17.48 


20.65 


Average 


16.16 


19.52 


18.08 


18.98 


Loss, Apr. 13 to Nov. 2 


7.28 


3.31 


1.86 


0.89 



percent and the wilting coefficient by the formula of Briggs and Shantz 
is 14.96 percent. Soil samples were taken to a depth of 6 feet for 
moisture and 3 feet for nitrate determinations. The data obtained iii 
1914 are summarized in Table 2. 

The bare-surface treatment sustained the least loss of soil moisture 
during the whole of this season. The 6-inch mulch was superior 
to the 3-inch mulch in checking loss by evaporation. The plots were 



Table 3. — Moisture content (percentage of dry weight of soil) of mulched and 
unmulched plots at Manhattan, Kans., on various dates in 1915. 



Date of sampling. 


Treatment. 


Weeds. 


3-inch mulch. 


6-inch mulch. 


Bare surface. 


April 15 


22.90 


23.00 


22.95 


23.00 


May 8 


23.50 


23.30 


23.14 


22,70 




23.90 


24.10 


23-94 


22.90 


July 5 


22.30 


23.80 


23.06 


22.80 


July 27 


21.60 


22.10 


21.81 


21.30 




21.80 


23.80 


23.00 


22.00 




21.70 


23.70 


22.90 


22.30 




22.50 


23.40 


22,90 


22.40 


Gain or loss, Apr. 15 to Sept. 8. . 


— 1.20 


0.70 


-0.05 


— 0.70 



CALL i'v' SKWKLL: THIC SOIL MULCH. 



55 



continued in 191 5 as in the previous season. Tlic data are summa- 
rized in Table 3. 

The precipitation for 1915 was very heavy, the total from January 
to September, inclusive, being 44.97 inches, as compared with 17.93 
inches and 29.89 inches for the same periods during 1914 and 1916. 
It is probable that the slightly higher moisture content of the mulched 
plots at the end of the season as compared with the bare surface plot 
is due to the heavy rainfall and the better condition of the surface 
to absorb water. 

The plots used in 1914 and 191 5 for the mulch studies were seeded 
to fall wheat in October, 191 5. In the spring of 1916 a new series 
of plots was established on the same soil type and handled in the same 
way. The results obtained are presented in Table 4. 



Table 4. — Moisture content (percentage of dry weight of soil) of plots vari- 
ously treated at Manhattan, Kans., in 1916. 



Date of sampling. 


Treatment. 


Weeds. 


3-inch mulch. 


6-inch mulch. 


Bare surface. 


May 29 


21.05 


21.16 


18.88 


20.25 


July I 


14-55 


15-52 


13.22 


15-99 


July 29 


1563 


19.91 


15-79 


19.07 


August 26 


14.08 


20.11 


16.91 


17-63 




14.40 


18.77 


17.69 


17-75 


Average 


15-94 


19.09 


16.49 


18.14 


Loss, May 29 to Sept. 23 ... . 


6.65 


2.39 


1. 19 


2.50 



The moisture content of the bare-surface plot exceeded that of the 
deep-mulch plot during the season and was about equal to that of 
the shallow-mulch plot. However, in considering the gain or loss 
of water on each plot, the difference in loss between the bare surface 
and deep mulch treatment was 1.31 percent in favor of the deep 
mulch. 

The experiments just reported did not account for the difference in 
the amount of moisture absorbed by differently treated plots, due to 
the condition of the surface soil. Accordingly, in 191 6, a series of 
plots similar to those in earlier experiments w^as laid out on which 
it was planned to retain all the water which fell upon them by means 
of earth dikes thrown up around their borders. Table 5 gives the 
data obtained from the diked plots. 

Comparing these data with the results obtained from the undiked 
series in 1916, it is noted that the bare-surface plot gained slightly 



56 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

more moisture than the deep-mulched or shallow-mulched plots where 
diked to insure equal absorption of rainfall upon all treatments. 



Table 5. — Moisture content (percentage of dry weight of soil) of diked plots 
variously treated at Manhattan, Kans., in 1916. 



Date of sampling. 


Treatment. 


Weeds. 


3-inch mulch. 


6-inch mulch. 


Bare surface. 




18.05 


20.20 


18.36 


16.77 


May 17 


20.42 


20.42 


20.02 


19.58 




20.48 


22.20 


20.5s 


20.04 


July 13 


12.65 


14.95 


13.30 


14-45 


August 12 


14.33 


19.40 


14.60 


16.40 


September 9 


14.40 


20.40 


20.73 


19.65 


Average 


16.72 


19.59 


17.92 


17.81 


Gain or loss, April 17 to Sept. 9 . 


-3.65 


0.20 


2.37 


2.88 



The average gain or loss by the different treatments is summarized 
in Table 6, the moisture content of bare mulched plots being used as 
a basis of calculation. 



Table 6. — Gain or loss of moisture from plots variously treated at Manhattan, 
Kans., in 1914, 1915, and igi6. 



Inches of water saved or lost by mulching. 



Treatment. 


1914. 


1915. 


19 

Undiked. 


16. 

Diked. 


Average. 


Deep mulch 


-0.37 


O.II 


-0.23 


— O.IO 


-0.130 




-0.48 


0.00 


— O.OI 


-0.50 


— 0.242 




-0.83 


— O.IO 


-0.75 


-1. 19 


-0.717 



EXPERIMENTS AT GARDEN CITY.13 

Experiments were begun at the Garden City station in 191 2 to 
determine the effect of cultivation on the loss of water by evapora- 
tion from the soil.^* The plots comprising this experiment were one- 
fortieth acre in size. They were irrigated each year during the winter 
as uniformly as possible in order to insure a high moisture content 
in the subsoil. 

During 1912 and 1913 the plots received three different treatments: 
(i) cultivation 6 inches deep; (2) cultivation 3 inches deep; and (3) 
no cultivation, but the surface kept bare of vegetation with a hoe. 
The plots were located on level land and were diked to prevent sur- 

13 The work at Garden City is in cooperation with the Office of Dry-Land 
Agriculture, U. S. Department of Agriculture. 
1* Lill, J. G. Report of the Garden City substation, 1914. In manuscript. 



CALL vV: SKWKLL: TIllC SOIL MULCIL 



57 



face run-off and drainage water coming onto ihcm. The soil was a 
silt loam with a deep subsoil. The water table at this elevation is 
located at about 75 feet. The subsoil was sufficiently porous to pre- 
vent saturation for any appreciable length of time. 

The results secured in 191 2 and 191 3 are presented in Table 7. 



Table 7. — Moisture content of irrigated mulched and unmulched plots at Garden 
City, Kans., in 1912 and 1913, expressed in total inches of water in 
the upper 6 feet of soil. 



Treatment. 


1912. 




July I. 


October lo. 


Gain or loss. 


May 9. 


Sept. 22. 


Gain or loss. 




17.89 


17-56 


-0.33 


18.52 


17.88 


— 0.64 


3-inch mulch 


16.76 


16.25 


-0.51 


17.41 


17.12 


— 0.29 


Bare surface 


17.15 


17-32 


o.'i7 


16.78 


16.58 


— 0.20 



Comparing the inches of water gained or lost by the three surface 
treatments, the bare surface has gained 0.5 inch more than the 6-inch 
mulch and 0.68 inch more than the 3-inch mulch in the 1912 season. 
In 1913, the bare surface gained 0.35 inch more than the 6-inch mulch 
and 0.09 inch more than the 3-inch mulch. 

In 1914 a series of dry-land plots was established. These were 
the same as the mulch plots established in 191 2 but received no irriga- 
tion. A fourth plot upon which weeds were allowed to grow was 
added to the series. The results obtained from both the irrigated 
and dry-land plots in 191 4 are presented in Table 8. 



Table 8. — Moisture content of irrigated and dry-land mulched and unmuched 
plots at Garden City, Kans., in 1914, expressed in total inches of 
water in the upper 6 feet of soil. 



Treatment. 


Irrigated plots. 


Dry-land plots. 


March 30. 


Sept. 16. 


Gain or loss. 


March 30. 


Sept. 16. 


Gain or loss. 


6-inch mulch .... 




17-59 


15-87 


— 1.72 


11.78 


12.35 


0.57 


3-inch mulch .... 




18.05 


16.62 


-1-43 


11.30 


II. 71 


0.41 


Bare surface 




17.76 


15.60 


— 2.16 


11.46 


11.96 


0.50 


Weeds 




16.41 


9.06 


-7-35 


10.75 


8.03 


— 2.72 



The weeds occasioned a considerable loss of water from the soil. 
On the irrigated plots, there was a loss of 1.72 inches from the 6-inch 
mulch, 1.43 inches from the 3-inch mulch, and 2.16 inches from 
the bare surface. This is a difference of 0.44 inch of water in favor 
of the deep mulch and 0.73 inch in favor of the shallow mulch. The 
dry-land 6-inch mulch and bare surface plots gained approximately 
equal amounts of water. The 3-inch mulch gained o.io inch less, 
while the weeds caused a loss of 2.72 inches. 



58 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



The average gain or loss with each treatment for each season is 
given in Table 9, data from the bare uncultivated plot being used as a 
basis for calculation. 



Table 9. — Average gain or loss of soil moisture from plots variously treated at 
Garden City, Kans., expressed in inches of water in the upper 6 
feet of soil. 



Treatment. 


Irrigated. 


Dry-land. 


Average. 


1913. 


1913- 


1914. 


1914. 


Deep mulch 


-0.50 
-0.68 


-0.35 
— 0.09 


0.44 
0.73 
-5-19 


0.07 
— 0.09 
-3.22 


-0.077 

— 0.032 

— 4.200 


Shallow mulch 


Weed s 





During these years at Garden City the precipitation averaged 17.07 
inches from January to January, being 18.74 inches in 1912, 23.58 
inches in 1913, and 9.7 inches in 1914. Evaporation from a free 
water surface averaged 53.06 inches. The wind velocity averages 
about 10 miles per hour, although velocities of almost 30 miles per 
hour prevail at times in the spring for 24-hour periods. 

It is evident from the data at hand, which were taken during seasons 
of severe climatic conditions, that a cultivated surface has not been 
more effective than a bare surface free of vegetation in preventing 
the loss of soil moisture by evaporation. They can lead to no other 
conclusion than that for Kansas conditions cultivation does not con- 
serve moisture except as it eliminates weeds and checks surface 
run-of¥. 

Effect on Development of Plant Food. 

The effect of cultivation on the development of plant food is often 
of more importance than the conservation of moisture. This phase 
of the problem was studied at Manhattan in 1914, 1915, and 1916, by 
determining the nitrates present in the differently treated plots previ- 
ously mentioned. The determinations were made to a depth of 3 feet. 
The data are presented in Tables 10 and 11. The results are sum- 
marized in Table 12. 

Table 10 shows that the bare-surface plot in 1914 contained the 
greatest amount of nitrates, with the 6-inch mulch plot second. This 
was a season of comparatively light precipitation (17.93 inches from 
January i to October i), but the rainfall was evenly distributed. 

The 1 91 5 season was one of heavy precipitation, 44.97 inches faUing 
from January i to October i. This came principally in the months 
of May, June, and July. For this season the nitrate development, as 
shown in Table 10, was again greatest in the bare-surface plot, but 



CALI. N: SKWKI.L: TIIK SOIL MUIXJI. 



59 



second in the plot having the 3-inch mulch. In such a rainy season, 
conditions less favorahlc to nitrification prevailed in the surface soil 
of the 6-inch mulch plot. 

Table 10— Founds of nitrates (NO^) per acre in the upper 3 feet of plots 
variously treated at Manhattan, Kans, in 1914 and 1915. 



RESULTS IN I914. 



Date of determination. 


Treatment. 












Weeds. 


3-inch mulch. 


6-inch mulch. 


Bare surface. 


Apri) 5 


264.38 


194.76 


328.29 


233.39 


May 19 


321.57 


332.47 


396-65 


281.06 


June 4 


179.61 


407.87 


579-34 


465.22 


July 9 


21.60 


583.37 


422.59 


464.53 


July 23 


32.84 


527.72 


878.86 


812.48 


August 5 


32.46 


618.67 


448.32 


1,128.15 


August 25 


41.57 


712.53 


456.79 


859.28 


October 9 


82.47 


424.21 


742.97 


1. 317. 32 


November 2 


142.87 


678.70 


697-55 


853.65 


Average 


124.37 


497.81 


550.15 


712.79 




-121. 51 


483.94 


369.26 


620.26 


RESULTS IN 191 5. 


April 15 


83 -43 


471.04 


276.83 


552.80 


May 8 


68.39 


386.60 


223.13 


325.98 


June 16 


43-92 


515.57 


247.40 


704.90 


July 5 


31.17 


555-55 


359-39 


709.22 


July 27 


32.38 


492.52 


328.71 


637.37 


August 24 


16.02 


398.61 


357-22 


599-41 


September 8 


19.87 


651.68 


484.26 


971-82 




42,10 


495.90 


325.20 


643.07 




-63.56 


180.64 


207.43 


419.02 



In 1 91 6, the months of July and August were very dry, although the 
precipitation from January i to October i amounted to 29.89 inches. 
In the diked plots, the average nitrates (Table 11) v^ere greatest for 
the 6-inch dry mulch, yet during the dry months of July and August 
• the nitrate contents of the bare-surface and 6-inch mulch plots were 
nearly equal. The average nitrate content of the plots receiving the 
bare-surface and the 3-inch mulch treatment were about equal. The 
data for the undiked mulch plots (Table 11) show the same results 
as those for the diked plots. 

The 6-inch mulch treatment is superior to the bare surface, and the 
bare surface and 3-inch mulch are about equal in nitrate development. 



6o JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Table ii. — Pounds of nitrates (NO3) per acre in the upper 3 feet of diked 
and undiked plots variously treated at Manhattan, Kans., in igi6. 



RESULTS FROM DIKED PLOTS. 



Date of determination. 


Treatment. 












Weeds. 


3-inch mulch. 


6-inch mulch. 


Bare surface. 




224.1 


287.7 


362.0 


302.9 




219.0 


174-3 


385-5 


251.8 


June 16 


102.5 


144.9 


274-1 


164.2 


July 13 


123-5 


372.7 


478.3 


465-9 


August 12 


100.2 


618.5 


617.7 


622.2 


September 9 


158.3 


531-5 


624.6 


445-7 


Average 


154.6 


354-9 


457-0 


375-4 


Gain or loss 


-65.8 


243-8 


262.6 


142.8 


RESULTS FROM UNDIKED PLOTS. 


May 29 


90.5 


68.2 


99-3 


115.8 


July I 


58.3 


246.5 


452.5 


327.9 


July 29 


67.3 


224.4 


256.2 


300.0 


August 26 


66.7 


343-0 


553-5 


424.0 


September 23 


109.7 


349-5 


584-0 


399-1 




78.5 


246.3 


S67-8 


313-3 


Gain or loss 


19.2 


281.3 


484-7 


283.3 



Table 12 presents annual and average data on the development of 
nitrates in the various plots from 1914 to 1916. 



Table 12. — Annual and average development of nitrates {pounds of NO3) per 
acre in the upper 3 feet of plots variously treated at Manhattan, Kans., 
1914 to 1916, inclusive. 



Year. 


Treatment. 


Weeds. 


3-inch mulch. 


6-inch mulch. 


Bare surface. 


I914 


124-3 


497.8 


550.1 


712.7 


1915 


42.1 


495-9 


325-2 


643.0 


1916* 


78.5 


246.3 


567-8 


313.3 




8T.6 


413-3 


481.0 


556.3 



^ Diked plats not included. 

The average development of nitrates for the 3 years has been greater 
in the bare, uncultivated plot than in the deep-mulch or shallow- 
mulch plots. 

Determinations were made of the total nitrogen contained in the 
weeds for the season of 191 6 in order to determine the amount of 
nitrification that had taken place in the weed plots. Converting the 



CALL iS: SEWELL: THE SOIL MULCH. 



6i 



total nitrogen into nitrates and adding it to the nitrates present in the 
upper 3 feet of soil of the weed plots gives the results shown in 
Table 13. This table indicates that as much nitrification took place in 
the weed plot as in the bare, uncultivated plot. 

Table 13. — Nitrification in zvecd plots, as indicated by the nitrates in the upper 
S feet of soil (in pounds of NO3 per acre) plus the nitrates in the 
weeds on September 23, igi6. 

Treatment. Undiked. Diked. 

Cultivated 3 inches deep 349-5 531-5 

Cultivated 6 inches deep 584.0 624.6 

Bare surface 399.1 445-7 

Weeds allowed to grow 348.2 474-3 

Summary. 

1. A cultivated soil is no more effective than a bare uncultivated 
soil in preventing evaporation. 

2. Cultivation conserves soil moisture by the ehmination of weeds 
and by preventing run-off. 

3. The development of nitrates may be as extensive without cultiva- 
tion as with cultivation. 



62 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



GREEN MANURING: A REVIEW OF THE AMERICAN 
EXPERIMENT STATION LITERATURE.^ 

A. J. PlETERS. 

Introduction. 

The use of clover or of some other legume in the rotation is gen- 
erally considered to be a cardinal agricultural practice in the humid 
sections of the United States. Without attempting to trace the de- 
velopment of this practice, which is an old one, it may be pointed out 
that in 1794 Thomas Cooper described a rotation of corn, wheat, 
clover two years, as being practiced by the best farmers in Pennsyl- 
vania. This is the most common rotation of today throughout the 
northeastern United States. The belief that clover was valuable as a 
soil improver rested first on experience and later, when the relation 
between the legumes and the nodule bacteria was discovered, men felt 
that the faith founded on experience had been justified by science. 

Many researches were made in which the tops and roots of clovers 
were analyzed, the quantities of nitrogen stored determined, and con- 
clusions drawn as to the degree to which the soil must be enriched by 
the growth on it of a vigorous crop of clover. 

Other legumes have come into use in the territory not well adapted 
to red clover, as the cowpea, Japan clover, and bur clover in the 
South, and crimson clover on the Atlantic Coast. The main function 
of all these is to maintain fertility or improve run-down soils. If 
these crops really fulfill this function the result should be larger crops 
following the legumes and the experience of farmers seems tO' warrant 
the conclusion that such is the case. It has been the practice of the 
American experiment stations to test many, if not all, of the common 
beliefs of agriculture to find how well founded agricultural practice 
is and how improvements might be made if needed. That the value 
of clover or other legume in the rotation or as a green manure would 
be so tested was to be expected. Therefore, it has seemed worth 
while to examine the literature of American experiment stations and 
to bring together our knowledge of this subject with such critical 
comment as may be warranted. 

^ Contribution from the Office of Forage-Crop Investigations, United States 
Department of Agriculture. Publication authorized by the Secretary of Agri- 
culture. Received for publication December 9, 1916. 



im1':ti:us: ckioi^n manukinc 



63 



The present iii([uiry concerns the evidence to he found in American 
experiment station hterature on the value of a legume as measured by 
the yields of succeeding crops. The writer trusts that this point of 
view will be constantly borne in mind, as it is not his purpose to 
criticize methods or results. It is not always possible, however, to 
agree with the authors of bulletins as to the conclusions drawn. 

In a general way a regional distribution of the literature may be 
made and this will at the same time be a distribution partially accord- 
ing to crops. In the South the rotation and green-manure work has 
been mostly with cowpeas, with a little on crimson clover, bur clover, 
sweet clover, and soybeans. In the North and Middle Atlantic States 
the experimental work has been nearly confined to crimson clover, 
while in the Northeast (including the Canadian Province of Ontario) 
and in the northern Mississippi Valley red clover has been the chief 
legume crop studied. Beyond the Missouri and in the Canadian 
Northwest, alfalfa, peas, tares, and species of Melilotus have been the 
chief leguminous green-manure crops. 

The experiment station bulletins will be examined by States accord- 
ing to the above rough geographical division. Many of the bulletins 
treating of rotations or of green manuring are popular expositions of 
the principles involved and contain no original experimental work. 
Such bulletins do not, therefore, come within the range of this study. 
The same is true of bulletins, many of them valuable, that treat of the 
relations between legumes and the nodule organism (B. radicicola) 
or report the results of the study of the amounts of nitrates present 
after various crops. While such studies may have an important in- 
direct bearing on the value of legumes they do not measure their value 
in terms of the succeeding crop and it is to this point of view that we 
shall mainly confine our attention. There are also bulletins and cir- 
culars from many stations containing the record of isolated observa- 
tions. In order not to consume too much space such bulletins will 
be mentioned only when considered important. A full bibliography 
of the subject is, however, on file in the Office of Forage-Crop In- 
vestigations and if any reader believes an important paper has been 
overlooked he will confer a favor on the writer by advising him. 

Experiments in the South. 

ALABAMA. 

The most important work with green manures in the South has 
been done in Alabama. 

Cotton. — It was early shown (Canebrake Bui. 7 for 1890) ^ that 

2 Complete references to all publications cited are given in the bibliography 
at the end of the paper. 



64 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



more cotton was produced on land immediately following cowpeas 
than on land on which peas had not been grown and that when the 
pea vines were turned under the yield of cotton was larger than when 
only the stubble was plowed under. The difference was not large, 
however, and the work appears not to have been repeated until several 
years later. 

The West Alabama station (Report, 1903, p. 2) showed that by 
rotating cowpeas and cotton the yield of cotton was on the whole 
increased over that secured by continuous cotton culture, though 
during one year the yield of seed cotton after peas was actually less 
than that after cotton. 

Further increases in the yield of both cotton and corn are reported 
as having been secured after peas, soybeans and sweet clover (Cane- 
brake Bui. 24). 

In Canebrake Bulletin 25 it is shown that the value of the cotton 
crop was increased by $8.68 per acre when a good stand of bur clover 
was turned under and by $7.08 per acre when a stand of crimson 
clover was used as green manure. 

Canebrake Bulletins 26 and 27 (1910) contain records of an ex- 
periment on the yield of cotton after various crops. Two years' yields 
are reported in Bulletin 27 and the table is reprinted here as Table i. 

Table i. — Effect of legumes and nonlegumes on cotton grown on the Houston 

clay. 



Cover crop in 1907 and 1908. 



Pounds per acre of seed cotton. 



1908, 



1909. 



Cowpeas in 1908, vines turned under 

Red top (fair stand), whole turned under 

White clover (good stand), whole turned under 

Red clover, hay cut 

Alfalfa, 17 months stand 

Common vetch, hay cut 

Common vetch and oats, hay 

Oats, hay 

Hairy vetch and oats, hay 

Hairy vetch, cut 

Crimson clover, cut 

Bur clover, cut 

No fertilizer, early cotton 1908-09 



656 
296 
440 
Hay 
760 
416 
351 
456 
752 
868 
640 
968 



1,204 
1.436 
1,464 
1.544 
1,512 
1.312 
1.436 
1,184 
1.384 
1. 512 
1.344 
1,360 
1,016 



While the total yield of cotton for the two years 1908 and 1909 was 
higher after bur clover, crimson clover, and hairy vetch than from the 
check plot, the difference was not very great. The plots on which 
hairy vetch and oats and common vetch and oats had been grown and 
cut for hay as well as the white clover plot yielded less than the check. 



rii;Ti;us: iiKi:i:N mani kinc 



65 



The total yield from the redtop plot exeeeded that from any legume 
plot save crimson clover and hairy vetch, hut it should of course he 
remembered that a crop of hay was taken from the lei^^ume plots and 
not from the redtop. While the turnin*;- under of le.j^ume stubble 
seems to have resulted in an increase of cotton, a careful study of the 
figures shows some peculiarities that make it desirable that such a 
test be repeated. The check plot gave the highest yield of cotton 
the first year but the lowest the second year. This might l)e expected 
because of the possible temporary depressing influence of a green- 
manure crop. However, the plot on which oats was grown in 1907 
gave the lowest yield of cotton in 1908 and the yield in 1909, though 
still lower than any plot having legumes turned under, was more 
than three times as heavy as it was in 1908. While some of the clover 
plots showed similar increases most of them did not. The red clover 
plot which yielded among the lowest in 1908 returned the largest crop 
in 1909, while the yield from the crimson clover plot in 1909 was the 
lowest but one from the legume plots, though in 1908 this plot had 
yielded more cotton than any other legume plot. 

A considerable increase in the yield of seed cotton following alfalfa 
is reported in Bulletin 26, page 14, but since the comparison of the 
yield after alfalfa is made with that which the field had produced in 
previous years the observation has only minor significance. 

The total yield of seed cotton grown for three years following the 
turning under of a crop of bur clover and one of crimson clover ex- 
ceeded the yield from a plot continuously in cotton by about 500 
pounds (Canebrake Bui. 27). It had been shown previously (Report, 
1905, p. 35) that when cowpeas were fertihzed and turned under the 
succeeding crop of seed cotton was increased by about 23 percent, 
but that on unfertilized land the cowpea crop was so poor that no 
increase in the cotton crop was noted. 

Many of the experiments showing the efifect of green manures on 
cotton have been brought together in Bulletin 120. In 1899 cotton 
followed cowpeas and velvet beans, the vines of which had been 
plowed under after the pods had been picked. The yields of seed 
cotton per acre were (p. 146) : 

Following cowpea vines i,533 pounds. 

Following velvet bean vines i,373 pounds. 

Following cotton 837 pounds. 

In another case (p. 151) cotton after velvet bean vines yielded 1,579 
pounds per acre; after velvet bean stubble, 1,126 pounds; and after 
cotton, 918 pounds. The average gain in cotton from plowing under 



66 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



the vines of summer legumes was 63 percent ; from legume stubble, 
18 percent. 

The effect of turning under crimson clover was to increase the yield 
of cotton, as is shown below (Bui. 147) : 



Corn. — The effect of a green manure on corn was studied as early 
as 1887 (Bui. 7), though there is no record of a check plot for that 
year. After two years of Melilotus, corn yielded 7 bushels more than 
on the check plot (Canebrake Bui. 24). The results of a number of 
experiments showing the effect of a green-manure crop on corn are 
recorded in Bulletin iii. In 1900 corn was grown on plots on which 
velvet bean stubble, second growth or entire crop, and the entire crop 
of beggarweed had been turned under. The yields are compared with 
those from unfertilized plots on similar soil 100 yards distant. It 
seems possible that the experimental plots were naturally more fertile, 
since the yield from the unfertilized field was but 5 bushels an acre, 
while that from the experimental plots ranged from 15.6 to 27.5 
bushels — a truly phenomenal increase if due entirely to the legume. 
Hairy vetch was compared with rye and turf oats as a crop to precede 
corn. The yields of corn in 1898 and 1899 were considerably better 
after the legume than after the nonlegume and a little better than on 
the plots left bare in 1897. Several peculiarities in the yields seem 
to point, however, to a lack of uniformity in the fertihty of the plots. 
For instance, on plot 6 hairy vetch stubble was turned under and the 
yield of corn was 16.8 bushels. On plot 11 the vetch was a failure 
and the stubble of what grew was plowed in. The corn yielded 18 
bushels on this plot. In 1899 plot 6 yielded 2 bushels per acre more 
than plot II. On the whole the vetch was clearly beneficial in 1898 
but the dift'erences in yields during 1899 on all plots were not very 
marked. 

In 1901 (Bui. 120 and 134) corn was grown on poor white sandy 
land, following corn or legumes. The yields were much higher after 
the legumes, reaching 81 percent increase on the plot on which velvet 
bean vines had been turned under. 

In the summing up Duggar gives the average increase in the yield 
of corn following a summer legume, vines plowed in, as 81 percent, 
and stubble plowed in, as 32 percent. 

Oats. — Red clover stubble increased the yield of oats by 15 bushels 



After oat stubble 

After crimson clover stubble 

After crimson clover, entire ripe and dry 



342 pounds seed cotton per acre. 
456 pounds seed cotton per acre. 



crop plowed under 



528 pounds seed cotton per acre. 



V 1 1 : r ! r s : ( ; k i -; i c n m a n u r i n ( ; . 



67 



per acre (Cancbrake Bill. 25), while after white clover the increase 
was 6.3 bushels. Oats after cowpeas turned under yielded 22.8 
bushels, while after millet the yield was 12.4 bushels (Bui. 95). In 
1897 oats were sown on six plots on which cowpea or velvet l)ean 
vines, velvet bean stubble, German millet, or weeds had been turned 
under. The average yield of oats from the nonleguminous plots was 
8.4 bushels; that from the legume plots, 32.6 bushels. It is admitted 
that these increases are probably greater than can ordinarily be ex- 
pected. (Bui. 95, 120, and 137.) 

In 1906 oats followed various crops. The yields following a 
legume w^ere in every case much higher than the yields following corn. 
The addition of 60 pounds of nitrate of soda per acre on a plot previ- 
ously in sorghum brought the yield of oats from that plot up to the 
average yield from all legume plots, but even the addition of 120 
pounds of nitrate of soda did not bring the yield from the sorghum 
plot up to that from the plot on which the entire crop of soybeans 
had been turned under (Bui. 137). 

Sorghum. — Sorghum for hay in 1897 yielded 85 and 86 percent 
more after cowpea or velvet bean vines turned under in 1896 than on 
a plot fallowed in 1896. 

In 1899 the following yields of sorghum hay were obtained after 
various legumes (Bui. 120) : 



The average increase in the yield of sorghum according to Duggar 
was 78 percent after summer legumes, vines turned under, and 57 
percent after stubble turned under (Bui. 120). 

Sorghum was also grown after rye stubble, crimson clover stubble, 
the entire crimson clover plant, and winter and spring weeds turned 
under. The yields were in every case much higher after the clover. 
In 1901, the average yield of two plots after rye stubble plowed under 
was at the rate of 5,992 pounds of green fodder, while the average 
yield after clover stubble plowed under was 11,230 pounds and the 
yield of one plot where the entire crimson clover plant was plowed 
under was 10,300 pounds. In 1903, sorghum after clover stubble 
yielded 13,000 pounds, while that after weeds yielded only 4,400 
pounds (Bui. 147). 



Yield per acre, 
tons. 



Increase from leg- 
umes, tons. 



After sorghum stubble 

After cowpea stubble .... 
After velvet bean stubble . 
After cowpea vines, picked 
After velvet bean vines . . 



3.65 
5.66 
5.80 
572 
6.76 



2.01 



2.15 
2.07 

311 



68 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Residual Effect of the Green Manure Crop. — An important feature 
of the Alabama work is the record of the residual effect of the green 
manure crop. In some cases this was clearly marked up to the third 
season. On the sorghum plots of 1899 referred to above, corn was 
planted in 1900. The yields from the plots on which cowpeas and 
velvet bean vines had been turned under in 1898 exceeded the yield 
from the plot in sorghum in 1898 by 3.6 and 2.6 bushels per acre, 
respectively. The yield from the cowpea stubble plot was also a little 
higher than that from the sorghum plot, but on the velvet bean stubble 
plot the yield was slightly lower than that from the sorghum plot 
(Bui. III). 

In 1900 corn followed cotton on plots on which velvet bean vines 
were turned under in 1898. The land was poor and the plot in cotton 
for two years before 1900 yielded 18 bushels, while from the adjoin- 
ing plot on which velvet bean vines had been turned under in 1898, 
25.5 bushels of corn were harvested (Bui. iii and 120). These plots 
in 1899 had yielded 1,578 pounds of seed cotton after velvet bean 
vines and 918 pounds after cotton, a difference of 660 pounds. In 
1900 the difference in the corn yield was 7.5 bushels on the same plots 
(Bui. 120). 

In 1901 corn followed corn of 1900 on legume plots (see p. 66). 
There is no record of a check plot here, but the yield from the plots 
'on which the entire velvet bean vines had been turned under exceeded 
that from the stubble plot by 59 percent (Bui. 120). Some plots in 
'cottoii in 1899 (Bui. 120, p. 146) were put into sorghum hay and 
'oats and sorghum hay as the next crops. A detailed financial state- 
ment (Bui. 120, p. 146) shows that the annual gain from turning 
umder a crop of cowpeas in 1898 was $14.32 per acre or a total for 
three years of $42.96 more than was realized under similar conditions 
but where no legume had been grown for many years. All crops w^re 
larger on the legume than on the nonlegume plot even to the third 
year after 'turning under the legume. 

Wheat and rye also showed marked increases after the turning 
under of a legume (Bui. 120). 

Relative Value of Turning Under the Stubble or the Entire Vine. 
— There are a number of experiments throwing light upon this ques- 
tion. With the exception of wheat and rye, crops were uniformly 
larger after the vines were plowed under than after the stubble only 
was saved. That the value of this increase, $5.98 per acre, was not, 
however, equal to the value of the hay crop sacrificed is pointed out 
in Bulletin 120, page 172. 

The record from this station shows that the yields of various field 



PIETERS: (JKKKN MANURING. 



69 



crops were larger following the turning under of the stubble or of the 
entire crop of a legume than after another field crop. The cumulative 
effect of the large number of observations and experiments, all agree- 
ing in the main, is unquestionably great. With the exception of red- 
top (Canebrake Bui. 27) and millet turned under (Bui. 95) no non- 
legume was used as a green manure unless the weeds mentioned in 
Bulletins 95 and 120 be so considered. The entire crop of redtop gave 
slightly more favorable results than the stubble of most of the legumes 
used. The turning under of a crop of millet was followed by a 
crop of oats but little more than half as larg^e as when a crop of 
cowpeas was turned under, 

ARKANSAS. 

Rotation experiments commenced in 1890 are mentioned in 
Arkansas Bulletins 18, 22, and 27 and yields on two plots are given. 
From these it appears that when cowpeas were turned under the yield 
of cotton in two years was larger than that obtained from the con- 
tinuous cotton plot in three years, and the first plot also yielded 22.4 
bushels of corn per acre as the second of the three crops following 
cowpeas. Corn grown on plots after legumes and after nonlegumi- 
nous field crops yielded larger crops after legumes. The average yield 
of five plots after cowpea stubble or peanuts was 36.8 bushels ; after 
the cowpea vines turned under, one plot, 39.7 bushels ; and after non- 
legumes, cotton, corn, etc., the average of seven plots was 25.47 bushels 
(Bui. 46). The increased yield after cowpeas plowed under did not 
pay for the hay sacrificed, but there is no record regarding the residual 
efifect of the cowpeas. Cotton also yielded 400 pounds more after a 
crop of cowpeas turned under than after cotton and 118 pounds more 
than after cowpea stubble (Bui. 46). 

Wheat was grown on plots after cowpeas, velvet beans, beggar weed, 
and soybeans, each alternate plot being cut for hay and the stubble 
only turned under; the entire crop on the other plots was turned 
under. The first and the tenth plots were left as checks, wheat fol- 
lowing wheat stubble. The turning under of the legume stubble in- 
creased the yield of wheat by 55 percent, but turning under the entire 
crop increased the yield over wheat stubble only 25 percent (Bui. 62). 
This depressing efifect of turning under an entire crop was also noted 
under certain conditions at the Alabama station. In another case 
(Bui. 58) the yields of corn and of cotton were larger after the legume 
vines had been turned under than after the stubble. Nothing is said 
as to the yield of hay. 

Oats followed various legumes turned under in 1900. Alternate 



yO JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

plots of oats were cut for hay and others for grain. The yields of 
both hay and grain were much higher after the legume than after oat 
stubble, the hay yield being nearly 60 percent and the grain nearly 
50 percent greater after velvet beans. There were no legume stubble 
plots. Oats following cowpeas, soybeans, and rye and vetch plowed 
in also yielded much more than following corn, sorghum, buckwheat, 
or oat stubble with 400 pounds of complete fertilizer (Bui. 66). The 
author points out that the legume crop turned under would have made 
hay worth more than the value of the increased crops. 

To test the relative value of cowpea stubble or the whole plant 
turned under seven plots, all in oats, were selected (Bui. 70). The 
average yield of the first three had been 21.84 bushels of oats; of 
the last four, 21. 11 bushels. On the first three oats followed oat 
stubble. On 4 and 6 oats followed cowpea stubble and on 5 and 7 
oats followed the cowpea vines turned under. The average yield of 
plots I, 2, and 3 was 24.38 bushels; of plots 4 and 6, 28.84 bushels; 
and of 5 and 7, 37.02 bushels. Cowpeas seeded in corn increased the 
yield of corn the following season.^ 

The residual efifect of cowpeas was studied between 1898 and 1903 
(Bui. 77). On four out of thirteen plots cowpeas were planted fol- 
lowing the 1898 wheat harvest and on two of them the whole plant 
was turned under, while only the stubble was used on the other two. 
Wheat was then grown for four years. The total yield on two check 
plots during the four years was 40.2 bushels ; on the two stubble plots, 
50.5 bushels ; and on the plots on which the whole cowpea crop was 
turned under, 56.6 bushels. In the last years of the series the yields 
were respectively 10, 11, and 12.8 bushels as an average of both plots. 
The stubble gave the best increase the first year, while the benefit 
from the whole plant was most marked the second year and was 
decidedly noticeable the third and four years. On two other plots 
cowpeas were planted after wheat each year after 1899 and the yield 
on these plots increased each year, reaching an average of 17.2 bushels 
in 1902. 

GEORGIA. 

This station furnishes but one bulletin of importance on this sub- 
ject (Bui. 27), but this contains the record of a remarkably clear 
and convincing experiment on the economic results of turning under 
a green crop or stubble. In 1893 and 1894 cotton was grown on 
plots in cowpeas the year before. On some plots the cowpeas were 
cut for hay, on others the peas were picked and the vines plowed 

3 Part of this information is repeated in the annual report for 1901. 



riKTKKS: (iKKKN MANUkINC;. 71 

[ 

under, while on still others the entire croj) of peas, vines, and pods 
was plowed in. Record was kept of the value of the crops during 
both years of each test with the results shown in Table 2. 



Table 2. — Yields of cotton after cowpeas variously treated, zvith combined 
values of the cotton and cowpea crops. 



Disposition of crop. 


Yield of seed cotton 
per acre, 1894. 


Total acre value of 

cotton, 1893, and 
hay and peas, 1892. 


Total acre value of 

cotton, 1894, and 
hay and peas, 1893. 




Pounds. 






Peas gathered when ripe 


2,004 


I50.61 


143-03 


Pea vines turned under green .... 


2,079 


42.96 


39.50 


Pea vines made into hay 


1,961 


55.89 


54.85 



The residual value of the veins has not been taken into account. 
The work of the Alabama station showed that this might be con- 
siderable but whether enough to pay for the hay sacrificed can not 
be told. 

MISSISSIPPI. 

The Mississippi station found that by rotating crops and fertilizing 
the productive capacity of the land was increased 17 percent in three 
years, while under continuous culture of one crop the productive 
capacity of the soil was decreased 16 percent (Bui. loi). Eight years 
after a system of rotation was started, the yield of cotton on continu- 
ously cropped but fertilized land was 252 pounds of seed cotton per 
acre, while on rotated and fertilized fields the yield was 404 pounds. 
The rotation included cowpeas as a catch crop in the corn. 

MISSOURI. 

During three years of cropping with cowpeas, peas one year in 
four, and also sown in the corn at the last working, the yield of corn 
was somewhat larger the second and third years on the plots having 
a catch crop of cowpeas than on the check, but during the first year 
the cowpeas slightly depressed the yield (Bui. 83). The differences 
were small, however, in each year and the authors state that the 
fertility of the plots was not uniform. The differences may there- 
fore have little significance. Wheat yields were less on the cowpea 
plots than on the checks in each of the two wheat years. The clover 
following wheat seemed benefited by the cowpeas in the second of the 
two years. On the whole the experiments did not lead to any definite 
conclusion. 

In experiments in southwestern Missouri (Bui. 84) some plots re- 
ceived phosphorus and potassium while the adjoining plots had cow- 



72 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

peas, phosphorus, and potassium. The yields of both corn and wheat 
were sHghtly less on the plots having the cowpeas. In an experiment 
in Jasper County (Bui. 119), cowpeas as a catch crop in corn did not 
increase the yields of corn, while the average yield of wheat was only 
very slightly increased. 

Bulletins 126, 127, 128, 129, and 130 describe soil experiments on 
various soil types ; plots receiving a cowpea catch crop in corn may be 
compared with the plots on which no legume was grown. In almost 
all cases the yields of corn were higher on the nonlegume plots, but 
those of wheat and oats averaged slightly better after the cowpea 
catch crop. 

In Bulletin 131, the statement is made that as a result of 12 years' 
observations it may be concluded that the removal of cowpeas, where 
wheat and peas are both grown each year on the same land, results 
in a gradual decrease of the wheat crop. No details are given. 

NORTH CAROLINA. 

In 1889 and 1890 this station raised wheat after cowpeas and after 
crabgrass. The plots were in duplicate and were repeated in 1890, 
making a very satisfactory experiment. Some plots were fertilized 
but in every case the crabgrass half of a plot received the same 
fertilizer as the cowpea half. On the unfertilized plots the yield of 
wheat on the cowpea half was double what it was on the crabgrass 
half, as an average of the two years. In 1890 the differences in yields 
were in some cases many times greater (Bui. 72, 77). It is said that 
the winter of 1889-90 was unusually severe and that winterkilling 
was especially bad on the plots that had not had cowpeas. 

TENNESSEE. 

In connection with a fertilizer experiment it was shown that, on the 
unfertilized plots, cowpeas turned under increased the yield of wheat 
over that from plots from which the cowpeas were removed (Bui. 
90). This was true in experiments on three farms. In Bulletin 96 
these data are repeated and yields of wheat also given from areas on 
which no cowpeas had been grown. On one farm areas with and 
without cowpeas can be compared only on plots fertilized with acid 
phosphate and potash ; the plots without cowpeas received a trifle more 
phosphate than the others. The yields of wheat were "much better 
after cowpeas had been turned under than where no cowpeas had 
been grown. On another farm wheat was grown on unfertilized 
plots on which cowpeas had been turned under or removed and on 



riKTKRS: GRKKN MANUKIN(J. 



73 



which no cowpcas had been e^rovvn. On the iinlinied portions the 
cowpeas made no appreciable difference in the yield but where lime 
was added yields were best after cowpeas turned under and next best 
on plots from which they had been removed. 

The value of cowpeas in buildin^^ up land is brought out in Bulletin 
I02. A poor soil was planted to cowpeas in 1910 and in 191 1. On 
one section, C, the crop was hogged off and on another, D, the vines 
were turned under for two years. In 191 2 corn was grown. Though 
the yield of corn on similar and adjoining land was less than 10 
bushels per acre that on the unfertilized plot of section C was 26.2 
bushels and on D, 31.9 bushels. It is not shown whether it paid to 
turn under these two crops of cowpeas but it seems very probable that 
hogging off was more profitable. 

In Bulletin 109 various statements are made but the present writer 
has not found the record of the evidence on which these are based. 
The conclusions appear, however, to be drawn from work done at 
the station and are as follows : 

1. Corn, sorghum and millet are not suitable crops to be grown for green 
manure. The yield of corn immediately following them was considerably re- 
duced. Rye is advised as a winter cover crop, but should be turned under early, 
when about i foot high, or less. 

2. The legumes can be advised as green-manure crops, but cowpeas and soy- 
beans when used for this purpose are not apt to be profitable for the first few 
years. If turned under each year and followed with wheat or other small 
grain, as can be done successfully at Jackson, the effect is cumulative ; that is, 
the area where the cowpeas are turned under gradually increases in productive- 
ness, while the area where either the peas are removed for hay or no peas are 
grown gradually becomes poorer until the difference between them is very 
marked. In the last two years of a 5-year trial there were obtained from 4 
to 9 bushels per acre more of wheat on the area where the cowpeas were turned 
under than where either none were grown or the crop was removed for hay. 

Sweet clover sown April 11, 1912, was cut once that year for hay. 
The crop was turned under May 13, 191 3, and followed with corn, 
which made a yield of 58.8 bushels per acre. On an adjoining plot 
where rye was turned under, the yield of corn was only 41. i bushels. 

SUMMARY. 

While it must be said that for the greater part the record above 
reviewed is that of a series of observations rather than of carefully 
planned and checked experiments it can not be denied that the ob- 
servations, all pointing in one direction, make a strong case for the 
value of leguminous green manure or rotation crops. The evidence 



74 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

will not, however, warrant more than this general statement. On 
important matters of detail, especially as to the relative value of dif- 
ferent legumes and the effect they may have on the succeeding crop, 
the record throws no light. For the most part this is because little 
work has been done with any plant save the cowpea. Only the 
Alabama station worked with winter legumes, and the value, if any, 
of nonlegumes for green manuring has practically not been touched 
upon. The question whether it pays to turn under an entire green 
legume crop has been taken up and it has been shown both by Ala- 
bama and Georgia that, considering only the first succeeding crop, 
the practice is not economical. Alabama showed, however, that 
cowpea and velvet bean vines turned under have a definite and con- 
siderable residual value, but the number of cases in which it is possible 
to compare the residual value of stubble or of vines are too few to 
warrant conclusions. This was also shown for cowpeas by Arkansas. 
This matter is important and it is to be hoped that experiments are 
under way, or will be undertaken, to determine this point definitely. 

The following statements are believed to be warranted by the 
experiments reported in the literature reviewed : 

1. The turning under of a leguminous crop or of the stubble and 
residues of such a crop results in an increase, sometimes large, in the 
next succeeding crop. 

2. Turning under a heavy crop of green plants, especially under 
conditions of dry weather or of imperfect working of the soil, may 
depress yields below those obtained from turning under the stubble 
only. 

3. Cowpeas grown as a catch crop in corn may depress the yield of 
the corn. This seems to be due to lack of moisture. 

4. The residual effect of whole crops of cowpeas and velvet beans 
turned under is marked, while that of the stubble is sometimes evi- 
dent, but slight. 

5. Considering only the effect on the first succeeding crop it is not 
profitable to turn under a full stand of green manure. 

Atlantic Coast Sfxtion. 
With the exception of a little work showing the influence of alsike 
and of red clover on potatoes, the work of the stations in this section 
has been almost entirely with crimson clover. One green manure 
crop of cowpeas was tried by the Maryland station and the New 
Jersey station records some work with cowpeas and soybeans. 



rii;'n-:Rs: (iRi:i-:N' m ,\ n inn nc;. 



75 



C ONNIX TK UT. 

The Connecticut (Storrs) station (Storrs Reports, icSc/j and 1900) 
grew potatoes in 1900 on land on which alsike clover had been seeded 
in July, 1899, and on land on which rye had been seeded after corn 
harvest. When turned under the clover was 3 inches and the rye 
3 feet high (Report, 1900, p. 63). The yield of potatoes on the 
clover plot, 183. i bushels of good potatoes, was considerably higher 
than that on the plot having rye and mineral fertilizers, 1514 bushels, 
but also exceeded that on the plot having clover plus mineral ferti- 
Hzers, 177.8 bushels. The difference in this case is not large but 
would appear to indicate a lack of uniformity in the fertility of the 
soil of the experimental area. Many details of the work are wanting 
and so far as known the experiment was not repeated. 

DELAWARE. 

The Delaware station has issued a number of bulletins on crimson 
clover, green manure, and orchard cover crops, but in only two cases 
is there any record of an experiment to determine the effect of 
crimson clover on a subsequent crop. A complete fertilizer including 
160 pounds of nitrate of soda per acre was appHed to one plot in 
June, 1890, while the crimson clover from $1 worth of seed was 
turned under on another plot in May. Sweet potatoes were planted 
on all plots. The yield on the clover plots was 18 bushels per acre 
more than on the fertilized plots and 103 bushels more than the 
average of five unfertilized plots (Bui. 11). In the report for 1892 
the yield of corn on a plot on which a heavy crop of crimson clover 
had been turned under is compared with that on adjacent land on 
which tomatoes had been grown. The corn on the latter plot received 
TOO pounds of nitrate of soda per acre, but still the yield was 18 
bushels less than that from the clover plot. So far as known this 
experiment was not repeated. 

MARYLAND. 

The Maryland station in Bulletins 31 and 38 records yields of 
potatoes after crimson clover. In one case a small increase, and in 
the other a 50 percent increase was found to result from turning under 
a green manure crop of crimson clover. In each case the results are 
for one year only. In 1895 the yield of corn on plots on which 
crimson clover had been turned under was 6.7 bushels per acre more 
than that on adjacent plots not having received crimson clover (Bui. 
46). The yield of corn on these plots in 1896 after a second crop 



76 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



of clover had been turned under was larger than the 1895 yield, but 
the yield of plots not having clover is not given for 1896, so it is not 
possible to say whether the added increase was due to season or to 
the extra clover. 

In an experiment on the availability of different forms of phos- 
phoric acid, plots were laid off receiving the same fertihzers but some 
of them having crimson clover seeded at the last working of the corn, 
others rye after the corn, and others left bare over winter (Bui. 68). 
The total yield of corn for 1895 on the crimson clover plots was less 
than that on the rye or on the bare plots, indicating that the fertility 
of the soil was less on these plots. In 1897, however, after two 
catch crops of clover and of rye had been turned under the yield of 
corn on the clover plots was 35 percent higher than on the bare plots 
and 58 percent higher than on the rye plots. In 1898, owing to early 
summer drought, wheat was sown instead of corn, the plots being 
fallowed for that season. There was no great difference in the wheat 
yields, those from the clover plots being slightly smaller. This work 
was continued through the crop of 1906 and the results reported in 
Bulletin 114. The results during the latter half of this experiment 
make it necessary to reverse the conclusions drawn from Bulletin 68, 
since the average yields on the rye plots during the years 1 902-1 906 
were better than on the crimson clover plots. The average yield of 
corn on the fallow plots and on those where crimson clover and rye 
were turned under were as follows : 

3 crops, 1895-1897. 3 crops, 1902-1906 

After crimson clover 45.9 bushels. 34.3 bushels. 

After fallow 45.0 bushels. 34.2 bushels. 

After rye 42.2 bushels. 40.3 bushels. 

This showing is really too favorable to the clover, since some of 
the clover plots had soluble phosphates while the others did not. The 
average yield of six crops of corn, two of wheat, and three of hay on 
all plots to which insoluble phosphates were applied were as follows : 



Table 3. — Average yields of corn, wheat, and hay after crimson clover, after 

rye, and after fallow. 





Corn, bus. 


Wheat, bus. 


Hay, pounds. 




39-4 


19.8 


3.500 


After fallow 


39-4 


20.8 


3,866 


After rye 


41.8 


23-7 


4,211 



The author suggests several reasons to explain the failure of clover 
to equal the rye, among them being the poor growth of the clover 
in recent years due to weather conditions. 



riirnoKs: c;ki:i:n manukinc. 



77 



Cowpoas were planted in 1896 and plowed under on limed and 
on unlinied plots. A rotation of wheat, hay, corn, followed. The 
wheat on the cowpea plots was much lar<^er than on the check plots 
(Hill, no, p. 9). 

MASSACHUSETTS. 

A special bulletin on green manuring by Dr. Julius Kuhn was issued 
in 1894, but the experiments discussed were all conducted in Germany. 

In certain other experiments potatoes were grown on clover sod, 
after one year of soybeans, and on land which had annually received 
a good dressing of nitrogenous fertilizer. All plots received annually 
equal applications of phosphoric acid, potash, and lime, but the clover 
land had received no nitrogenous fertilizer for 16 years. The yield 
of potatoes on the clover sod was almost as large as that on the 
fertilized plots and besides two cuttings of clover hay had been taken 
from this land. Soybean stubble did not improve the yield of pota- 
toes and the yields of oats and rye, v^hich were grown for some years 
alternately with soybeans, steadily declined (13th and i6th Ann. Rpts.. 
pp. 94 and 123, respectively). Soybeans did not improve the fertility 
of the soil as measured by the yield of oats in an experiment sum- 
marized in the 9th Annual Report, pp. 176-177. 

NEW JERSEY. 

At the New Jersey station (Reports, 1912 and 1913) corn or oats 
were grown on land on which green manure catch crops of rye and 
of crimson clover or vetch had been turned under each year from 
1908 to 1912. The yields of corn for 1909, 191 1, 1912, and 1913, 
and of oats for 1910 were in every case conspicuously larger after 
crimson clover or vetch than after rye. 

An experiment on using cowpeas as a summer catch crop w^ith 
rye and with wheat was carried on for five years, 1909-1913 (Report, 
1912, p. 261, and Report, 1913, p. 473). The wheat crop of 1909 
before any legume had been used v^as 19.6 bushels on the plot with- 
out legume and 31.07 bushels on the legume plot, while the average 
yield of wheat for the four following years was 11.87 
bushels respectively. While it is true that during each of these years 
the yield of wheat on the legume plot was higher than that on the 
nonlegume the difference was scarcely more than that which existed 
between the plots when the experiment began. The yields on these 
plots for 1914 are given in Bulletin 281 and it appears that the legume 
plots continue to return the largest yields. 

On the rye plots the effect of the legume is more marked. If, 
however, the crop of 1910 instead of the one for 1909 is chosen as a 



78 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

basis for comparison, the effect of the legume on the wheat crop be- 
comes more evident. The crops for 1910 were preceded by a very 
poor stand of cowpeas, and on the legume plots the yield of wheat 
was but little more than that on the nonlegume plot, while the yield 
of rye was even less on the legume than on the nonlegume plot. In 
every subsequent year the crops on the legume plots were markedly 
higher than on the nonlegume plots. 

A comparison of the amount of dry matter in a corn crop following 
crimson clover stubble or the entire crop turned under is made in the 
Report for 1894, p. 135. The yield of crude fat, fiber, protein, ash, 
and carbohydrates in the corn was greater where the entire clover 
crop had been turned under, but the total food secured from the land, 
including clover and corn, was greater when only the stubble was 
turned under. 

In the 35th Annual Report, pages 223-226 (1914), Lipman and 
others report that the yield of rye was nearly a fourth more after 
legumes than on plots without legumes and that the yield of wheat 
was more than doubled. The percentage of nitrogen recovered was 
also higher in the grain crops from the legume plots. 

In the course of an experiment to determine whether ground lime- 
stone aids in the decomposition of organic matter, green alfalfa, 
timothy, oats, and peas were chopped fine and added to soil in pots 
(Report, 1914, p. 217). To other pots nitrate of soda was added, 
while the check pots as well as all others received equal amounts of 
mineral fertilizer. Buckwheat was used as the indicator crop. On 
the pot to which alfalfa had been added the amount of dry matter 
produced was nearly as great as on the pot to which the nitrate of 
soda was added, while the amount of dry matter on the timothy and 
on the oats and pea pots was greater than that from the check pots 
but did not equal that from the alfalfa pots. 

This station has also conducted a number of cylinder experiments 
in which leguminous crops were grown and turned under for a suc- 
ceeding crop. The yields of rye and corn were larger after crimson 
clover and hairy vetch had been turned under in the soil on which 
these legumes w^ere grown, than on soil on which legumes had not 
been grown but to which an equivalent amount of green legumes had 
been added (Bui. 250). More nitrogen was also recovered in the 
corn crop in the former case. Further cylinder experiments are re- 
ported in Bulletins 288 and 289. Oats yielded better in 1909 and 
19 10 after hairy vetch than in cylinders without legumes (Bui. 288). 



rii/n:Rs: (iKi:i:\ m \.\ukin(J. 



79 



Crop rotations were coiulucted in 3J0 cylinders (Bill. 289) on one 
series of which i^^reen manure catch crops were used according to the 
followini:;- plan: Corn followed by crimson clover; potatoes followed 
by cowpeas and vetch; oats with cowpeas or soybeans between oats 
and rye, which was again followed by cowpeas or vetch. Another 
series of cylinders received stable manure at the rate of 15 tons per 
acre every two years ; a third series received nitrate of soda at the 
rate of 160 pounds per acre. All received lime and mineral ferti- 
lizers and of two other series one received only lime and a second lime 
and mineral fertilizers. 

The amount of dry matter in the crops on the green manure series 
was much larger than that from any of the other series. Studies 
were also made of the amounts of nitrogen removed in crops and the 
amount left in the soil at the conclusion of the experiments. While 
the amount of nitrogen in the soil at the end (1912) was less than at 
the beginning (1907) in almost all cases this loss was least in the 
green manure series and from this series the largest amount of 
nitrogen had been removed in the crops (pp. 29-31). 

While the effect of the association of a legume and nonlegume, 
strictly speaking, is not included in the present study, its bearing upon 
the value of a legume is evident and a brief statement of the litera- 
ture on the subject will be included therein. Since the most important 
single contribution has been made by the New Jersey station, the 
entire literature will be reviewed here. 

In 191 1 Lyon and Bizzell (Cornell Bulletin 294) gave analyses 
of timothy growing with and without alfalfa and of oats growing with 
and without Canada field peas. It was shown that the nonleguminous 
crop grown with the legumes contained a markedly higher protein 
content than that grown without. 

In Bulletin 253 of the New Jersey station (1912), Doctor Lipman 
reviews the older literature on the subject, criticizes the method 
adopted by Lyon and Bizzell, and reports extensive experiments 
made by the New Jersey station between 1908 and 191 1. These ex- 
periments consisted of growing legumes and nonlegumes in cylinders, 
in pots in the greenhouse, and in some cases in field plots, in which 
case the nonlegume was grown both alone and in association with a 
legume. To determine whether or not nitrates pass by diffusion from 
the legume to the nonlegume, Doctor Lipman used small pots, in some 
cases glazed and in others unglazed, sunk in the soil contained in a 
larger pot, the legume being grown in the one and the nonlegume in 
the other. As a result of his experiments Doctor Lipman concludes 
that under favorable conditions nonlegumes associated with legumes 



8o JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

may secure large amounts of nitrogen from the latter, even though 
this may not be indicated by an increased proportion of nitrogen in 
the dry matter of the nonlegume. In the field experiments no satis- 
factory results were obtained. In some cases the yield of the non- 
legume was depressed because of the presence of the legume. This 
is ascribed to lack of moisture. 

A part of this experimental work had already been reported by 
Lipman (Jour. Agr. Sc., 3 : 297) and a further discussion in regard 
to the methods and the priority of the work was entered into by 
Doctors Lyon and Bizzell and Doctor Lipman in the Journal of the 
American Society of Agronomy (Vol. 5, No. 2, pp. 65-82. 1913). In 
Bulletin 253 of the New Jersey station Doctor Lipman quotes from Bul- 
letin 61 of the South Carolina station to show that the beneficial effect 
of peas on the yield of corn has been noted by that station. The present 
writer has, however, been unable to find that the statement made on 
page 8 of Bulletin 61 of the South Carolina station is justified by the 
yields reported. It is stated that the yield of corn was increased, 
particularly on plot 4. Examination of the yields, however, shows 
that during 1898 the yield on plot 4 was less than 0.4 bushel more 
than that of the average on all the plots in the experiment, and in 
1899 the yield on plot 4 was actually less than the average of the 
other plots. 

Westgate and Oakley (Jour. Amer. Soc. Agron., 6:210-215. 1914) 
report the analyses of a number of grasses growing with and without a 
legume. They conclude that the data presented are not sufficient to 
w^arrant the statement that a nonlegume growing with the legume will 
have an increased content of protein. 

At the Virginia station an experiment was conducted on the effect 
of the association of legumes and nonlegumes (Va. Technical Bui. 
I, April, 191 5). The authors show that bluegrass growing with 
white clover and timothy growing with red clover did not contain 
more protein than when growing without, except that in the second 
year of the experiment the timothy growing with clover contained 
more protein than when growing without clover, but this may well 
have been the effect of the preceding year's growth of clover. Corn, 
when grown with beans, produced a greatly increased crop and the 
percentage of nitrogen was also much larger when the corn was grown 
with beans than without. In the field experiment, however, com 
grown with beans gave a lower yield than when grown without. The 
authors consider that the decreased yield in this case was due to the 
lack of moisture. 

Evans (Jour. Amer. Soc. Agron., 8:348-357. 1916) has recently 
shown that bluegrass and timothy growing with clovers have a higher 



riETERS: GREEN MANURING. 



8l 



l^rotcin content than when growing alone on check plots and also that 
the growth of grasses made between clippings when grown with clover 
exceeded that of grass grown alone by some 21 percent. 

RHODE ISLAND. 

Beginning with 1898 crimson clover was seeded between certain 
rows of blackberries and of blackcaps at the Rhode Island station, 
while other rows were left blank (Bui. 91). The crops of berries in 
1901 and 1902 were about twice as large on the rows receiving 
crimson clover as on the check rows. 

In 1896 an experiment was started on growing corn continuously 
with clover and with rye as green manures. The total yield of hard 
corn for nine years, 1897 and 1905, inclusive, on the clover plot was 
378 bushels per acre and on the rye plot, 312.2 bushels. Beginning 
with 1898 the plots were divided so that some remained fallow. The 
yields on the fallowed sections averaged a little less than on the rye 
and much less than on the clover. From 1898 to 1902 there was a 
progressive improvement on the clover plot, as measured by the corn 
yield (Bui. 113). 

A series of rotations was planned in 1893 (Ann. Rpt., 1893, p. 
176), in one of which a legume replaced timothy and redtop, rota- 
tion B ; while rotations E and F were alike except that rye was used 
as a cover crop in F, and a legume was used in E. 

In the annual report for 1897 it is shown that the yield of potatoes 
following clover sod on rotations A and C exceeded that following 
corn on rotations B, D, E, and F by nearly 55 bushels. This increase 
consisted, however, almost entirely of small potatoes, the yield of 
large ones being only bushels more after clover sod than after 
corn. Potatoes followed grass and clover sod on rotation C and 
followed corn on rotation D. In Bulletins 74 and 75 the yields of 
potatoes for the first and second courses of these rotations are re- 
ported. The yield on C (grass and clover sod) was less than i bushel 
per acre greater for the two courses than from D (after corn, follow- 
ing grass and clover sod). 

Rotations E and F are discussed in Bulletin 76, in the Annual Re- 
port for 1902, and in Bulletin 167. In the latter the yields for the 
entire period of 20 years have been brought together. These rota- 
tions were alike save that clover was seeded with the timothy in E 
and not in F, and that legumes, crimson clover and hairy vetch, were 
seeded at the last working of the corn on rotation E while rye was 
used on F. 

The average yield of corn and potatoes on all plots of these two 



82 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



series for 20 years, extracted from tables on pages 14 and 19 of Bul- 
letin 167, are as follows : 

Rotation E, Rotation F, 

bushels. bushels. 

Corn (hard) 56 53 

Potatoes 119 216 

This shows no benefit, under the conditions obtaining at the Rhode 
Island station, from using a legume rather than grass or rye. Atten- 
tion must be called, however, to the fact that on the continuous corn 
acre a leguminous cover crop gave better returns than rye (Bui. 113). 



SUMMARY. 

Potatoes and corn have been the chief indicator crops of this sec- 
tion, with wheat and oats of very minor importance. Crimson clover 
has been the chief green manure crop, red and alsike clover, cowpeas 
and vetch having been occasionally used. The records include care- 
fully controlled experiments and observations of small value, but the 
results are in general agreement as showing that a legume benefits the 
following crop of corn or potatoes, especially the latter. 

The addition of nitrogenous fertilizers did not always bring the 
yield on the nonlegume plot up to that from the legume plot. 

Rye was compared with clover as a green manure crop to precede 
potatoes and corn by the Connecticut, Maryland, and New Jersey 
stations and the clover was generally found to be superior. 

The entire clover crop turned under was found by the New Jersey 
station to yield a crop of corn containing more protein, fat, and 
carbohydrates than a corn crop on clover stubble, but the total amount 
of these feeding stuffs obtained in the two years was much greater 
when the clover was used for hay. 

{To he continued in the next number.) 

A LIMESTONE TESTER.^ 

Cyril G. Hopkins. 

The writer has designed a simple apparatus, by means of which the 
relative purity of limestone can be quickly ascertained with a very 
satisfactory degree of accuracy, the determination being based upon 
the basicity, or carbon dioxid content. As shown in figure 4, this 
limestone tester consists of two small glass bottles, joined together 

1 Contribution from the Department of Agronomy and Chemistry, College 
of Agriculture, University of IlHnois. Received for publication November 18, 
1916. 



HOPKINS: A I-IMHSTONI-: TKSTKR. 



«3 



and fitted with groimd-^lass stop])crs, the stopper of the smaller bottle 
resting upon a surface only sliii^htly inclined from the horizontal, and 
projecting loosely into the neck, thus serving as a valve. 

To make the test, place 5 grams of pulverized limestone in the 
larger bottle and fill the smaller one to the side opening with acid 
made by mixing about equal parts of concentrated hydrochloric acid 
and water and saturating with carbon dioxid. Insert the stoppers 
and weigh. Now tip the apparatus carefully until the acid begins 
to flow through the side opening. As it drops upon the limestone, 
the carbonate is changed to 




chlorid and the liberated car- 
bon dioxid gas passes through 
the side opening, lifting the 
small stopper as it passes out. 
Partly immerse the apparatus 
in cool water to keep it at 
about room temperature. 
Gradually transfer the acid 
until foaming ceases : then dry 
the apparatus with a soft cloth, 
weigh, and note the loss. To 
this loss in weight add about 
0.6 milligram for each cubic 
centimeter of air space in the 
loaded apparatus (see Tables 
.2 and 3), then deduct the 

proper percentage for the _ 

room temperature (about i ^ ^ . . ,. 

^ ^ Fig. 4; Apparatus used m testing limestone, 

percent for 20° C. see Table weight, about 40 grams; capacities below con- 
l), and divide by 2.2 to get necting tube, about 35 and 40 cc, respectively 

the relative purity of the stone. 

If the direct loss represented only the total carbon dioxid liberated, 
then its weight divided by 2.2 would give the relative purity of the 
stone in terms of calcium carbonate, since 5 grams of pure calcium 
carbonate contains 2.2 grams of carbon dioxid. 

However, the moist air which fills the air space at the beginning is 
replaced by moist carbon dioxid during the reaction. At 20° C. and 
745 millimeters barometric pressure (taken as room temperature and 
average atmospheric pressure at an elevation of 600 feet above sea 
level), I cubic centimeter contains 1.754 milligrams of carbon dioxid 
or 1. 1 54 milligrams of air, not including the water vapor. The 



84 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

difference in weight is 0.6 milligram per cubic centimeter ; and, if the 
air space in the loaded apparatus is, for example, 75 cubic centi- 
meters, then, under these conditions, 45 milligrams should be added 
to the loss in weight. 

Again, the gas (air or carbon dioxid) passing out of the apparatus 
during the reaction is accompanied by some water vapor, which 
amounts to .017 milligram per cubic centimeter at 20° C. The com- 
bined weight of the carbon dioxid and water vapor in i cubic centi- 
meter, at 20° C. and 745 mm., is 1.771 milligrams ( 1.754 -|- .017) . 
Thus, under these conditions, about i percent must be deducted from 
the first corrected weight. 

For example, 5 grams of a certain limestone shows a loss of 1.99 
grams. The first correction (45 milligrams) increases this to 2.035 
grams, and the second correction (20 miligrams) reduces it to 2.015 
grams. This divided by 2.2 gives 0.916, or 91.6 percent, as the rela- 
tive purity of the stone. 

For most practical purposes, the first correction is a constant for 
each apparatus, the variations for ordinary differences in temperature 
and pressure being negligible ; and the second correction varies signif- 
icantly only with change of temperature. Thus, if the room temper- 
ature is 30° C, add to the weight of escaped gas 0.6 milligram per 
cubic centimeter of air space and then deduct 1.74 percent (see Table 
i). This, in the above example, with a direct loss of 1.99 grams, 
would give a final corrected weight of 2 grams of carbon dioxid from 
5 grams of stone, and this divided by 2.2 gives 90.9 percent. But, 
to perform the operation at 30° and compute the second correction at 
I percent, as should be done for 20°, would introduce an error of 
0.7 percent in the purity found. 

To saturate the hydrochloric acid with carbon dioxid, drop a piece 
of limestone weighing 3 or 4 grams into a 500 c.c. bottle of the 
diluted acid, replacing the stopper after foaming ceases. To de- 
termine the air space in the loaded apparatus, place 5 grams of pulver- 
ized limestone in the larger bottle, fill the smaller bottle to the side 
opening with water, and then pour in measured water from a gradu- 
ated cylinder and note the addition required to completely fill the 
apparatus. The vapor pressure of the dilute hydrochloric acid used 
is negligible, corresponding to less than i milligram of HCl per liter. 

A set of weights from 5 milligrams to 50 grams, a balance suitable 
for these weights with a capacity of 100 grams, a thermometer, a 25 
c.c. graduated cylinder, and the limestone tester, are all one needs for 
determining the relative purity of limestone for use in neutralizing 
acidity, as in soil improvement. If one also has a barometer and a 
balance capable of weighing to i milligram, a still higher degree of 



HOPKINS: A LIMESTONE TESTER. 



«5 



accuracy may bo .secured b}' using^ tbc data ^'wcn in the accoini)any- 
ing tables. 

Thus, 5 grams of a pulverized limestone shows a direct loss of 
2.164 grams at 25° C. and 724 mm., with an aj)paratus having 71 
cubic centimeters air space when loaded. When saturated with water 
vapor under those conditions, i cubic centimeter contains 1.659 milli- 
grams of carbon dioxid or 1.093 milligrams of air, the difference 
being 0.566 milligrams, or 40 milligrams in 71 cubic centimeters. Hiis 
first correction being added gives 2.204 grams of moist carbon dioxid, 
of which 1.29 percent, or 28 milligrams, is water vapor, leaving 2.176 
milligrams of dry carbon dioxid, and this divided by 2.2 gives 98.9 
percent; whereas, if the barometric pressure were assumed to be 
745 mm., the purity found would be 99.0 percent, as may readily be 
computed from the data given in Tables i, 2, and 3. (Tables i and 
3 are computed from basic data given in Castell-Evans' Physico- 
Chemical Tables, I, 341 ; and Table 2 is Parr's table extended to 
include the higher summer temperatures.) 



Table i. — Carbon dioxid saturated with water vapor at 760 millimeters 

(29.92 inches). 



Temperature. 


-Milligrams per cubic centimeter. 


Percentage of 
water in total. 


Pressure of 
water vapor, 
mm. 


°C. 


°F. 


Carbon dioxid. 


^\'ater vapor. 


Total. 


10 ^ 


50.0 


1.879 


0.009 


1.888 


0.50 


9.2 


II 


51.8 


1.870 


.010 


1.880 


.53 


9.8 


12 


53.6 


1.862 


.011 


1.872 


• 57 


10.5 


13 


55-4 


1.853 


.011 


1.864 


.61 


II. 2 


14 


57-2 


1.844 


.012 


1.856 


.65 


II.9 


15 


59-0 


1.836 


.013 


1.848 


.69 


12.7 


16 


60.8 


1.827 


.013 


1.840 


.73 


13.5 




62.6 


1. 818 


.014 


1.832 


.78 


14.4 


18 


64.4 


1.809 


.015 


1.824 


.83 


15.4 


19 


66.2 


1.800 


.016 


1. 816 


.89 


16.3 


20 


68.0 


1. 791 


.017 


1.808 


■95 


17.4 


21 


69.8 


1.782 


.018 


1.800 


1. 01 


18.5 


22 


71.6 


1.773 


.019 


1.792 


1.08 


19.7 


23 


73-4 


1.763 


.020 


1.784 


1. 15 


20.9 


24 


75-2 


1.754 


.022 


1.776 


1.22 


22.2 


25 


77.0 


1.744 


.023 


1.767 


1.29 


23.6 


26 


78.8 


1-735 


.024 


1-759 


1-37 


25.0 


27 


80.6 


1.725 


.026 


1.751 


1.46 


26.5 


28 


82.4 


1-715 


.027 


1.742 


1-55 


28.1 


29 


84.2 


1.705 


.029 


1.734 


1.64 


29.8 


30 


86.0 


1.695 


.030 


1.725 


1-74 


31.5 


3^ 


87.8 


1.685 


.032 


1. 716 


1.85 


33-4 


32 


89.6 


1.674 


-033 


1.707 


1.96 


35.4 


33 


91.4 


1.664 


.035 


1.699 


2.08 


37.4 


34 


93-2 


1.653 


.037 


1.690 


2.20 


39.6 


35 


950 


1.643 


•039 


1.682 


2.34 


41.8 



86 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



00 \0 O 
0\ M ro Lo t4 
\0 r- i> 



00 o N (3 
00 00 O 0\ 0\ 



0\ O <N CO to O 
CO ro M o 0\ 
00 00 00 00 00 



00 On O M M 
00 t>- \0 



ro ^ O \0 
ro M o 
i>- r- 



r- 00 GO 

0\ 00 O 

o o o o o 



00 00 t> 1> vO 
fO N M O 

\0 O O vO >o 



vo On O M 
ro (N M o O 0\ 
oo 00 00 00 00 r~ 



fO 00 
00 \0 

r- 



O O O M 
ro ro CN M O 
x> 



fO ro 

0\ 00 O 

o o o o o 



ro ro ro (N N 
ro M M o 
vO O O O \0 



O N lo \0 
<N <M H O OsOO 

00 00 00 00 



00 0\ O N ro 
O \0 lo •Tj- 
r~ 



lO <5 O 
ro IN M O 0\ 
O 



00 00 00 00 

00 vO 

vO nO O nO o 



00 00 00 I> 
ro IN M O 
O O O O 



U 



IN M O 0\ 0\ 00 
00 00 00 i> 



ro O t-~ 00 
r- 



O 



ro "^t 
00 r- vO 
O O O o 



N 

fO N M O 0\ 

o o o o ^ 



M (N ■rt lO !>• 
M M o 00 
00 00 00 



oo 0\ w M ro 
O lO ro 
i>- t- 



\0 00 

M o a 00 

o O 



oo O O On 
O <jO ro 
vO O vO O "O 



OV ON 00 00 00 

N M O On 00 
>0 'O O <-0 i-O 



mD On O in 
M o 0\ 00 00 
00 00 



ro vO 00 
O ro 01 

1^ 



On O M (N 
M M O 0\ 00 

t~~ o o 



ro ^ 
vO o o o o 



•rf ro ro ro 
N h-t o ON 00 
O nO nO lO »0 



0\ M M lO !>• 

O O On 00 O 
00 00 I-- 



00 O M M ro 
lO VO M 
i> 



lO O O 00 

M o On 00 
X>- nO nO O 



ON ON On O O 
O ^ •rf ro 
■O nO O O nO 



O O On On On 
(N HH On 00 

"O O lO lO lO 



^ O t-~ On O 
O On 00 NO 
00 r-~ 



ro VO -O t- On 
to ro M M 
t— r~ i>- 



O M M ro r/^ 
M O On 00 

sO nO nO 



to lO lO 
vO to Tj- ro 01 
nO nO nO nO vO 



to to to in lo 

M O ON 00 
NO nO to to to 



On M oi to 
On On 00 r~ O to 
r~ r-~ 



00 O M ro 

ro IN M 
r- 



to O 00 On 
O On 00 O 
O nO O "O 



On O O O O 
to to ^ ro 01 
O O O nO o 



O O O O On 
M o On 00 nO 
■O O to to to 



Tj- NO On M N 
On 00 O NO to 
i>- 



to O oo On 
Tj- ro 01 M o 



O M Ol fO 

O On 00 I> vO 
r^.NO nO nO nO 



to t0\0 NO 

to ro 01 M 

O nO nO nO nO 



NO nO nO nO to 
O On oo \0 
\0 to to to to 



On i-i N Tto 
00 00 O to 
!>. 



On O 01 ro ^ 
ro 1^ (N M o 
r- 



lO nO !>• 00 On 
On 00 \0 to 
\0 nO nO nO nO 



O o M M M 

to -rf ro oq M 

nO nO ^O nO 



O l> On i-i oq 
00 \0 to to 
t~- 



tJ- lO J> 00 On 
ro N M o On 
!>• sO 



O oq ro Tt 
On 00 !>• nO to 

nO nO nO nO nO 



to nO nO \0 
ro 01 h-i O 

NO NO nO nO nO 



i> NO no 
On 00 l> nO to 
to to to to to 



On M N tJ-nO 

!>. t— nO to 

t-~ 



On O 0< ro ^ 
01 01 M O On 

nO 



O 00 On O 
00 \0 to to 
\0 nO nO nO nO 



O M hH 01 (N 

ro oq M o 
nO nO nO nO O 



0^ 01 01 01 M 
On 00 NO to 
to to to to to 



OiHOirO'^to nOI^ooOnO 



HOPKINS: A LIMKSTONI-: I KSTFCK. 



00 vO M O 



rj- lo 00 O O 
O O 00 \0 O 

0\ 00 00 00 00 00 



M N fO >0 
»0 Tj- PO N M 

00 00 00 00 00 



O 00 00 

O O 00 o 

00 



00 00 0\ 00 00 
ir) 't PO (N M 
I- 



00 o >o 

O 0\ 00 1^ O 
r» O O O O 



O O N m rf XT) 

0\ O 00 r- o ^ 

00 00 00 00 00 00 



O 00 0\ o 
ro M M M 

00 00 00 00 00 



M N cs ro ro 

O O 00 vO 

00 



ro N M O 
O 0\ 00 O 
\0 O O O 



Tt ir> 00 0\ o 

0\ 00 vO »0 VD 

00 00 00 00 00 00 



M ro rt rf- »o 

ro N M O 

00 00 00 00 00 



vO 00 00 On 
O 00 r~ O 
r~ r~ r-- 



0\ 0\ 0\ Ov 0\ 
Tt ro C") M o 



0\ 00 vO O 
0\ 00 \0 >o 
\C o o o o 



O O M ro vo 
00 00 o m Tj- 

00 00 00 00 00 00 



00 0\ O M 
ro c^) M M o 

00 00 00 00 00 



M N fO PC 

O 00 r~ vO 
t~~ r- 



^ ^ 

rj- ro <N w O 



ro PD N M 
C\ 00 \0 »o 

\0 o *o o o 



Tj- in 00 0\ O 
00 o lo 

00 00 00 OO 00 00 



N PO o 
PO P^ 1-1 O 0\ 

00 00 00 00 



r~- 00 0\ O 
00 O >o ^ 



o o o o 

PO PO M O 

r- 



0\ O 00 00 r~ 
00 O 'i" 
\0 O O O O 



O O P^ PO o 
po 

00 00 00 00 00 00 



00 0\ O M 
p>j M o O 0\ 
00 00 00 00 r- 



IN PO PO 

00 sO 
r- t-- 



1_0 1_0 LO UO lO 

po M o 0\ 
i> O 



Tj- PO PT) 

00 O 

o o o o o 



IT) 00 O M 
-O po po 

00 00 00 00 00 00 



N PO lOvO 
N M O 0\ 00 

00 00 00 i> 



00 0\ 0\ o 
i>- 



o o o o o 

PC P<1 M o o\ 

vO 



O 0\ On 00 00 
00 O lO Tt PO 
^3 



OONP^-^O t-oOONOM WPO'^Tl-u-) lOOvOOO lO'^^POPO 

vO'Oi'^'^POf^ mOO\O\00 r^\0>0'^PO P^mOOsOO t^sO^O'^PO 

00 00 00 00 00 00 oooor^t^r^ r^r^r^t^r^ r^t^r-OO 

Tfuor^OOOM WPO'^lOO r-OOOvOO mmmmm 000\00\ 

\OiO"^POPOM MOOsoOt^ O'^'i"'^P0 P>^i-iOC\00 r^vO'^POP^ 

000000000000 oooor^t-i:^ t^r-r-r^t^ r-r^i>-00 oOvOOO 



Ov O <N PO u-5 vO 

U-> Tt PO P^ M 

00 00 00 00 00 00 



^~ 00 O M M 
O On 0\ 00 

00 i:^ i>- 



PO PO ir> to 
O to ^ PO P^ 

r-- i> 



"O "O nO 
M o On 00 
vO o o 



NO O to 

NO to tJ- PO o< 

nO nO nO nO nO 



tJ- to 00 O M 
to PO N C<l M 
00 00 00 00 00 00 



p^ po to \0 r~ 
o On 00 O 
00 i> r- 



00 On On O M 
to PO PO P^ 
r- 



M P^ (N 
O On 00 

\0 O NO 



<N M M O O 

\0 PO <N 

\0 nO O nO nO 



On O CM PO too 
^ PO M O 
00 00 00 00 00 00 



t~ 0\ O M CM 

On 00 00 NO 



PO rj- to to o 
to PO CS M 

t>. 



NO r~ t-- 
O On 00 r~- o 
\0 \0 nO O 



1> NO vO 

to PO CM M 

\Q \0 O nO o 



^ to 1> 00 O M 

rj- po P^ H M O 

00 00 00 00 00 OO 



<M ■<:1- to O 
On 00 !>• \0 to 
!>• I— 



00 On O M M 

■rf PO PO CM M 

t-- 



CM M M CM 

O O 00 O 
nO nO vO nO 



P^ M O O 
to Tj- PO CM M 

\0 nO O nO o 



P>» PO to 



nO 00 On O 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



fa 


OOOOtJ-MO OOO'^MO OOvOTfMO OOO'^MO 000^0)0 
OiHroiJ0i>O\ OfN^Ooo 0\M(r)u-)i— oooc^-^vO t^O\i-iroLo 
loioiouoiolo vooooo ot^r^t^r^ t^oooooooo oooooOsOs 






OMvooioo ■^0\roooM t-MuoO'^ oocMOOro r^O'^r^o 
OOOnOnoooo r-vOOiOio •^'^roooc^ mm00\0\ oooot^OO 


0\ 
0\ 




roooror^nr^ moouoO tToomom loosrooo -^r^M-^r^ 
OOsOsoooor^ r^vOOiOTf rtrorocN<N mooOnO oot^r-oio 


6\ 

M 


o 


OioO\Tt-0\fO ooroi>WvO Mioa\rOoO (NOOroi:^ M'^oOMTt 
OOQOoor^r^ vOOioio^ ^trocMtNn mooooo ocr-ooio 




00 


0\0oooor-x:— OOiJ^'^'st- roroNCNM OOOOsoo r-r^oi^io 


O 
O 

d 




roooroooiNt^ NOwvOO ^ONfOr^M OOrj-ooM lOOMioO 
OsOOOOt^t^O \OlOlO->^Tj- (-o(>4(MiHM OOCAOOOO r^ooiorj- 


00 

a\ 

00 
01 




Oioovoo-s}- OrooONt— mvOO^On rot^Mirjoo cqvOO\ro-0 
OsOOOOt^OO u-i rt m rOOJCNMO OOsOOOr-' t^voiou-i'^ 


o 

0\ 
00 




r^Nt^MOM vOOioOs'* ooror^Mvo O'^oofNu^ O\f000ro 
oooor^t^sOO loio'^roro <n(nmmo oO\oooot^ OOioiorj- 


00 
00 





■^O^^^oo^OOO <nI>mvOm uoO'^oON t^MioO\ro ^COror^O 
oot^t^OOi^ lO'^'^rorO (NCSmOO O\O\00f-t^ vOOuo-^'* 


00 


00 


OiooioO'^ O\rfoorooo cst^MioOs •rtooc^r^o ror^MTj-t^ 
oor^J>'OvOio Tf'^rorotN nmmoOn O\ooooi>r^ O^oiO'd-ro 


00 




t^(Nt-~Ml>M Omij-)0'^ 0\fOOOCM\0 Mir>0\ror^ O'^oOMr}- 
r-t^vOOuovo rj-Tj-rorOM mmooOn Osoot^r^o Oio^t-^ro 


00 

to 

00 
<N 




TtO^Osrooo roooNt~-M OOiOO\fO OONOO^ oOmiooon 
r^vOvOiJ^io-^ rj-forocq(N mmOO\0\ oOQOr-t-O uoio-^roro 


o 
to 

00 
(N 


M 


MOOiooio O^On'^oo ror^NOO loOsror^M ir)oocsu-)0\ 
r^vOOioiO'^ '^rocNtNn mOOC?\0\ 00I>-i>>OvO u-!-^->^r<:i<N 

MMMMMM MMMMM MMMOO OOOOO OOCOO 


00 





\n \Ci \/~> IT) Tf rorOMMi-i mOOnOnoo oor— t^ouo iOTtrcro(N 


ro 

00 

01 


s / 

/ ^ 
/ ° 


OMMrO'^iO \0 <Xi Os O M<Nro"^iO Ot^00C>O mnoO'^io 


m 
<u 
Xi 

u 
a 



HOPKINS: A LIMESTONE 



TESTER. 



89 





00 'S" N 


00 ^ N 


00 CM 


00 tJ- c^i 


00 't n 







6 M CO >o 0\ 
10 10 10 to »o to 


d CM 00 

vO vO 


0\ 1-H ro >0 
I- I- 


0C3 d r^ -"t 

00 00 00 00 


1-^ d\ >-< ro 10 

00 00 0\ On 0\ 





r- 


<^ 00 N M 
(N CM 

N W N M 


10 00 ro 
M M O 0\ 


M 10 0\ ro 
00 00 


M Tt 00 H 

10 10 ro ro 


>0 00 M 

N M M C> 


M 

ro 

d 

ro 


00 


10 rt 0\ (T) 00 
ro M N M 

N N N CM 


CM M to 

M 0\ 


00 CM vO 
00 vO 


00 M 10 00 
10 -rj- ro 01 


to 00 M -rj- 

CI 0\ 


"i- 

d 
ro 




CM t-~ M to 

rt ro rO CM CM M 
(N CM N CM CM CM 


0\ ro 00 CM vO 
0\ 0\ 00 


H to On ro r~ 

00 NO to 


M to 00 CM to 

to ^ ro ro CM 


OV CM to 00 M 
M M 0\ Oi 


vO 

d 

ro 




0\ enoo fo M 

rO M CM M M 
0» CM CM CM CM CM 


\0 to On ro 
0\ 00 00 


00 CM 

I> to 


00 CM to On CM 
ro 01 (N 


On CM 0\ 
M On 00 


00 

q 
d 












ro 




>o 00 
ro CM M M 

CM C>) CM CM (N CM 


ro r-- CM 
Os 0\ 00 00 


to On ro r~ M 
sO nO to to 


to 0\ M NO 
ro ro <N CM 


roO roO 
M 0\ 00 




d 

ro 





N r~ CM M to 
ro CM CM M M 
M CM CM <M CM <N 


0\ ro r- 
0\ 00 00 


M T(- 00 

\0 to Tt- 


(N nO Oi ro 
ro CM 0) i-i 


ro ro 
M On On 00 


OJ 

0( 
& 

CM 


00 


^ 00 fO 00 CM 
CM <N M M 
CM CM CM CM CM CM 


M to 

Os On 00 00 


00 M M 10 
nO to 10 


0\ roo ^ 
ro ro 05 CM M 


^ 
00 00 


00 

6\ 


VO 
10 


to 0\ 

M CM M M 0\ 
CM CM CM CM CM M 


ro 00 CM r- M 
0\ 00 00 !>• 


to 0\ 00 CM 

to to 


NO ro M 
ro ro <N iH M 


00 M 
On On 00 r- 
M 


d\ 

CM 




CO t-~ CM r~ M 

CM M M 0\ 

CM CM CM (N M M 


to 0\ ro 00 
0\ 00 vO 


(N i-i to On 
to to rj- ro 


ro 00 
ro CM M 


M to 00 M T}- 

0\ 00 00 

M 


00 

NO 

6\ 

CN 


N 
tri 


0\ 0\ ro 00 ro 
M M On C\ 


M >o to 
00 00 i> vO 


On ro 00 CM vO 

to to Tj- CO 


M 10 

ro CM iH M 


00 CM to 00 r^ 
On On 00 
00000 


On 

CM 



to 


M M 0\ 00 


'^t 00 ro N 
00 


NO On ro 
to to Tl" ro ro 


M ^ 00 CM 
CM CM M 


to 0\ CM 0\ 
0\ 00 00 !>• 
00000 


ro 
to 

d 

CM 


00 


ro 00 CM t-» CM 
M 0\ 0\ 00 


M 10 rj- 0\ 
00 i> to 


ro !>• M vO 
to ro ro 


rt" 00 M 10 On 
M M M On 


ro NO 0\ ro 
00 
00000 


to 

ds 

CM 


\o 


rj- 0, rj- 00 ro 
M 0\ C\ 00 00 


00 CM I> M 
t^o vO to 


00 ro 
to •Tj- ro ro CM 


M to 00 CM nO 

CM M On 


ro vO ro 
0\ 00 mD 
00000 


ro 

d\ 

CM 


E / 














g / 

/ ^ 

/ ° 


M ro to 


J> 00 Ov 


M M ro to 

CNl CM C^J CM CM 


NO 00 0\ 

CM CM CM CM ro 


M (M ro ^ to 
ro fO ro ro ro 


CO 

u 



go JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



AGRONOMIC AFFAIRS. 
NEW BOOKS. 

Productive Farm Crops. By E. G. Montgomery, Professor of Farm 
Crops, New York State College of Agriculture, Cornell University. 
J. B. Lippincott Co. (Philadelphia and London), 1916. 21^ cm. 
Pages 501 -j-xix; figs, 203; col. frontispiece. 

There has been a great need for a comprehensive textbook in one 
volume covering the entire subject of economic farm crops. The de- 
mand comes largely from secondary agricultural schools and from 
short-course students in agricultural colleges. Students of college 
grade are perhaps better taught by the lecture method. Such a 
book should be complete so that the student who is educated in agri- 
culture may know something of the crops grown outside the region 
where he lives. It should be written in a language which may be un- 
derstood by student who is without technical training, but must not 
be so simple as to be uninteresting. 

The space allotted to each crop should bear some proportion to the 
importance of that crop to the whole production. It should be so ar- 
ranged with appropriate headings and subheadings that it is teachable. 
" Productive Farm Crops " seems to meet all of these requirements 
in a satisfactory manner. Many other good books have been written 
which cover the same subject. In fact, practically all of the infor- 
mation contained in " Productive Farm Crops " may be found in other 
well-known works, but these other books usually are written for some 
particular region, are limited to a certain crop or group of crops, or 
else they contain a mass of irrelevant matter as soil physics, orchard 
management, and stock judging. 

Professor Montgomery has produced a first-class text book. 
Where but one volume is desired it would be valuable also as a ref- 
erence book. For reference purposes, however, a collection of two 
or more of the leading agricultural books, as one on the grains, an- 
other on forage crops, and perhaps another on fiber crops, would be 
more serviceable. 

It is customary in reviewing a book to find some fault with it. In 
this case it may be suggested that sugar-cane growing is of enough 
importance in this country to merit discussion ; that sugar beets 
should be treated apart from mangels ; and that hemp, hops, chufas, 
cassava each deserve a short paragraph. Sunflowers and artichokes 
might also be mentioned. These omissions are not serious and sink 



AtJKONUMU AKKAIRS. 



9' 



into insignificance when compared with the merit which the l)()C)k pos- 
sesses. — Lyman Carrier. 

MEMBERSHIP CHANGES. 

The membership reported in the ])revious issue was 609. Since 
that time 2 members have resigned and 11 new members have been 
added, making- a net gain of 9 and a total membership at this time of 
618. The names and addresses of the new members, the names of 
the members resigned, and such changes of address as have come to 
the notice of the Secretary are printed below. 

New Members. 

Clark, Geo. H., Seed Commissioner, Dept. Agr., Ottawa, Canada. 

Criswell, Judson H., 2807 Quarry Road, Washington, D. C. 

FiNNELL, H. Howard, 203 Boys' Dormitory, College of Agr., Stillwater, Okla. 

HoDSON, Edgar A., 207 Delaware Ave., Ithaca, N. Y. 

Holland, B. B., Box No. 46, Route i, Memphis, Texas. 

Kennedy, P. B., 11 Budd Hall, College of Agr., Berkeley, Cal. 

LiPPiTT, W. D., Great Western Sugar Co., 500 Sugar Bldg., Denver, Colo. 

Miller, Frank R., Bowker Fert. Co., 43 Chatham St., Boston, Mass. 

Mathews, Oscar R., Belle Fourche Expt. Farm, Newell, S. Dak. 

Ratliffe, George T., Belle Fourche Expt. Farm, Newell, S. Dak. 

Winters, N. E., Supt, Substation No. 3, Angleton, Texas. 

Members Resigned. 

Laidlaw, C. M., Physics Dept., Ontario Agr. College, Guelph, Ontario. 
Stoddart, Chas. W., Dept. Chem., Pa. State College, State College, Pa. 

Addresses Changed. 

Atwater, C. G., Agr. Dept., The Barrett Co., 17 Battery Place, New York, N. Y. 

Bliss, S. W., 17 Fifteenth Ave., Columbus, Ohio. 

Du BuissoN, J. P., Senekal, OFSP., South Africa. 

Kemp, W. B., Maryland State College, College Park, Md. 

McLane, J. W., Bur. Plant In^us., U. S. Dept. Agr., Washington, D. C. 

Tucker, Geo. M., R. F. D. No. 8, Chevy Chase, Md. 

Schmitz, Nickolas, Pennsylvania State College, State College, Pa. 



92 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



NOTES AND NEWS. 

Victor L. Cory, superintendent of the Denton (Texas) substation 
at Krum, Texas, has resigned and on March i sailed for Freetown, 
Sierre Leone. 

James W. Day has been appointed assistant in agronomy and 
Joseph R. Neller, research assistant in soils, at the New Jersey 
station. 

R. J. H. De Loach, for the past several years director of the 
Georgia station, has resigned and is now in charge of the service 
bureau of the Armour Fertilizer Co., with headquarters at Chicago. 

C. A. Dorchester, assistant in farm crops, and J. A. Krall, of the 
farm crops extension force, have exchanged positions in the Iowa 
college for the college year 191 6-17. 

A. R. Evans, instructor in farm crops in the University of Mis- 
souri, has resigned to accept a position in the Office of Markets of 
the U. S. Department of Agriculture. 

Edward T. Fairchild, president of the New Hampshire college since 
1912, died at Durham, N. H., on January 23. 

J. N. Harper, director of the South Carolina station since 1905, has 
resigned to take charge of the service bureau of the Southern Fer- 
tilizer Association, with headquarters at Atlanta. J. C. Pridmore, 
associate professor of agronomy at the University of Tennessee, re- 
signed on February i to become associated with him in this work. 

Arthur Huisken is now an assistant in soils at the Ohio station. 

Ove F. Jensen, who was engaged in graduate study at Iowa State 
College last year, since July i has been assistant in crop production at 
that institution. 

Clarence C. Logan, assistant in soil extension, Ward H. Sachs, as- 
sociate in chemistry, and F. C. Richey, assistant in soil physics, all of 
the Illinois station, have resigned. 

M. L. Nichols has been appointed assistant professor of agronomy 
and assistant agronomist in the Delaware college and station. 

M. E. Olson, formerly farm crops superintendent at the Iowa 
station, on September i resigned to enter the Office of Corn Investiga- 
tions, U. S. Department of Agriculture. 



ACKONOM U AI' l' AIKS. 



93 



A. M. Peter is actinia director of the Kentucky station, not direc- 
tor, as reported in the January Journal. 

D. W. Pittman is now instructor in agronomy and assistant agrono- 
mist in the Utah college and station. 

George Roberts, treasurer of the American Society of Agronomy 
and head of the department of agronomy of the Kentucky College 
of Agriculture, has been appointed acting dean of that college, vice 
J. H. Kastle, deceased. 

W. H. Mclntire, assistant in soil chemistry at the Tennessee sta- 
tion, has recently returned from leave of absence for graduate study 
at Cornell University, where he was granted the degree of Ph.D. 

G. L. Schuster became assistant in farm crops and R. G. Wiggans 
assistant professor of farm crops at Ohio State University with the 
beginning of the college year. 

John B. Smith has been appointed assistant in crops at the Mis- 
souri station. 

T. H. Stafford has succeeded H. L. Joslyn as assistant professor 
of soils at the North Carolina college. 

A. M. Ten Eyck, agricultural agent in Winnebago County, 111., 
since 1914 and previous to that time extensions professor of soils at 
the Iowa college and professor of agronomy at the Kansas college, is 
now in charge of the agricultural service bureau of the Emerson- 
Brantingham Company, manufacturers of agricultural implements, 
with headquarters at Rockford, 111. 

R. S. Thomas as assistant in soils and W. R. M. Scott as assistant 
in farm crops are recent appointments at Purdue University. 

At the annual meeting of the American Association for the Ad- 
vancement of Science, held in New York, December 26-30, the 
address of Dean Eugene Davenport, the retiring vice-president of 
Section M, agriculture, was entitled " The Outlook for Agriculture." 
The topic discussed at the meeting of this section was " The Adjust- 
ment of Science to Practice in Agriculture." Different phases of 
this topic were discussed by Dr. H. J. Wheeler, of Boston, Dr. G. F. 
Warren, of Cornell University, Director J. G. Lipman, of the New- 
Jersey station, and Director B. Youngblood, of the Texas station. 

The International Dry-Farming Congress, the National Irrigation 
Congress, and the International Soil-Products Exposition were held 
at El Paso, Texas, October 19-26, 191 6. President W. M. Jardine, 



94 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



of the Dry-Farming Congress, outlined the history of the congress, 
told of the progress of the dry-farming movement, and stated that 
the work of the future must include the encouragement of livestock 
production, the betterment of home conditions, and the provision of 
profitable work for the farmer and his family throughout the year. 
Other speakers were Director Youngblood and A. H. Leidigh of the 
Texas station. President E. G. Peterson of the Utah Agricultural 
College, Director Forbes of the Arizona station. Professor Throck- 
morton of the Kansas college, Col. A. J. Bester of South Africa, 
Niel Nielson of Austraha, and J. M. Romagny of France. Gov. 
Frank M. Byrne of South Dakota was elected president for 1917. 

Meeting of Southern Agronomists. 

The eighteenth annual convention of the Association of Southern 
Agricultural Workers was held in New Orleans, La., January 24-26, 
1917. The officers of the association for the past year were W. M. 
Riggs (S. C), president; T. D. Boyd (La.), vice-president; and Dan 
T. Gray (N. C), secretary. General sessions were held in the morn- 
ings, while the afternoons were devoted to special sessions of the 
livestock and agronomy sections. H. A. Morgan (Tenn.), was chair- 
man of the agronomy section, and T. E. Keitt (S. C), secretary. 

Perhaps the most important feature of the meeting from an agro- 
nomic standpoint was the report of the committee on coordinating 
investigational work in agronomy in the South. This committee con- 
sisted of C. B. Williams (N. C), chairman, and C. A. Mooers 
(Tenn.), A. F. Kidder (La.), C. K. McClelland (Ga.), and T. E. 
Keitt (S. C). After presenting a list of projects in agronomy now 
under way at the Southern stations and showing wherein there was 
an overlapping in these projects, the committee made certain recom- 
mendations which may be briefly outlined as follows : 

1. That eight main soil provinces be recognized and that these provinces be 
used as a basis in adjusting agronomic experimentation among the States. 

2. That new work be planned to eliminate duplication, particularly on soils 
of the same character. 

3. That work now in force be adjusted so as to correlate and eliminate 
duplication. 

4. That certain important fundamental data be reported in connection with 
each experiment. 

5. That the States Relations Service be requested to designate a man to col- 
lect and classify the projects in agronomy at the Southern stations and to advise 
the station directors regarding work being done at other stations in the South. 

6. That duplication of work in plant production be confined, so far as pos- 
sible, to duplication on different soil types or under different climatic conditions. 



ACiKONOMU AKl-AIRS. 



95 



7. That in varietal tests the source of seed be j^nvcn and be uniform, if 
possible. 

8. Tliat varietal test's be conducted mainly to determine the best strains for 
local conditions and that improvement work be started with these best strains 
after they are determined. 

9. That the agronomic work at the stations be in charge of men trained and 
educated along agronomic lines. 

Local Sections. 

At the fall meetings of the Iowa section, the progress of experimental work 
in agronomy in various States has been discussed. Officers of this section 
elected in October are: Clyde McKee, president ; H.W. Warner, vice-president; 
Ove F. Jensen, secretary; and R. H. Bancroft, treasurer. 

The North Carolina local section of the American Society of Agronomy was 
organized at West Raleigh on December 19, 1916, The following officers were 
elected: C. B. Williams, president; C. L. Newman, vice-president; and W. F, 
Pate, secretary-treasurer. It is probable that the section will have about 20 
members when organization is completed. 

The thirteenth regular meeting of the Washington (D. C.) section was held 
at the Cosmos Club, December 18, 1916. The annual report of the secretary- 
treasurer was presented, after which the following officers were elected for the 
ensuing year : C. E. Leighty, president; John S. Cole, vice-president ; A. C. 
Dillman, secretary-treasurer ; and P. V. Cardon and Chas. E. Chambliss, addi- 
tional members of the executive committee. Mr. C. H. Clark then presented 
an illustrated paper on " Experiments with Seed Flax as a Winter Crop for 
the Southwest," in which it was pointed out that a large part of the world's 
flax crop was produced from fall seeding and that preliminary experiments in 
our own Southwest indicated the possibility of the establishment of an impor- 
tant new center of flaxseed production there. Mr. Frank C. Miles followed 
with an illustrated paper entitled " Production of Hemp and Flax Fiber in the 
United States." Mr. Miles gave statistics of the production of these two fiber 
crops, showed where and how they are produced, and outlined and illustrated 
the methods by which the fiber is separated from the woody portions of the 
stalks. 

The fourteenth regular meeting of the Washington (D. C.) section was held 
at the Cosmos Club on January 17, 1917, at 8 p.m. The general subject for 
the evening, " Field Stations of the Bureau of Plant Industry," was discussed 
by several speakers, with illustrations. " The History and Development of 
Field Stations " was outhned by Dr. B. T. Galloway, after which the work of 
the field stations of the various offices was discussed by the following : plant 
introduction gardens, P. H. Dorsett; dry-land stations, John S. Cole; irriga- 
tion stations, C. S. Scofield ; forage-crop stations, H. N. Vinall ; and cereal 
stations, H. V. Harlan and Chas. E. Chambliss. The program was followed by 
a social hour, with refreshments. 



96 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

DIRECTORY OF LOCAL SECTIONS. 

Georgia State College. 

President, 

Secretary, Geo. A. Crabb. 

Cornell University. 
President, B. D. Wilson. 

Iowa State College. 

President, Clyde McKee. 
Secretary, Ove F. Jenson. 

Kansas State Agricultural College. 
Secretary, C. C. Cunningham. 

Minnesota College of Agriculture. 

President, C. P. Bull. 
Secretary, C. H. Bailey. 

North Carolina. 

President, C. B. Williams, West Raleigh. 
Secretary, W. F. Pate, West Raleigh. 

Ohio State University. 

President, 
Secretary, 

South Dakota State College. 

President, A. N. Hume. 
Secretary, Manley Champlin. 

New England. 

President, 

Secretary, Earl Jones, Amherst, Mass. 

Washington, D. C. 

President, C. E. Leighty. 
Secretary, A. C. Dillman. 



JOURNAL 

OF THE 

American Society of Agronomy 



Vol. 9. March, 1917. No. 3. 



LIVESTOCK AND THE MAINTENANCE OF ORGANIC MATTER 

IN THE SOIL.i 

Elmer O. Fippin. 

Organic matter in the soil is universally recognized as necessary to 
its largest productiveness. Maintenance of the organic matter con- 
stitutes one of the most trying problems of practical farming. In 
this paper it is desired, first, to press the importance of the organic 
constituents perhaps a little further than has commonly been done ; 
and second, to direct attention to the effect of animals on the organic 
matter in the feed consumed in a way that has not been emphasized 
in the ordinary discussion of soil maintenance and livestock 
husbandry. 

Physical and Nutrient Functions of Organic Matter. 

All of us readily concede the important physical eft'ects of organic 
matter in its partially decayed form, in respect to structure and tilth, 
moisture capacity, color, and heat absorption. We also emphasize 
its importance as a source of moderately available plant nutrients and 
the preeminent storehouse of combined soil nitrogen. It is our 
custom to say that the plant substance is broken down in the process 
of decay and the nutrients reduced to very simple and soluble forms, 
such as nitric acid and the ions of the bases. These may be taken 
up by the roots of the growing plant and used in building new 
structures. 

It is not long since the use of nitrogen was limited to the nitrate 
form. The destruction of the highly organized nitrogenous sub- 

1 Presented at the ninth annual meeting of the American Society of Agron- 
omy, Washington, D. C, November 14, 1916. 

97 



98 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



stances of organic material, the liberation of ammonia, and its oxida- 
tion to nitrates by the various steps involving the intervention of 
microorganisms has been explained as the limiting factor in the 
availability of organic carriers of nitrogen. 

Without minimizing the operation of those processes, it is now 
known that many plants use ammonia and the amino radical as a 
source of nitrogen. Some plants (e. g., rice)^ seem to prefer the 
ammonia form of nitrogen to the nitrate form. An experienced 
English florist and gardener at Auburn, N. Y., states that it is pos- 
sible to force flowers and fruits in the greenhouse by exposing vats 
of liquid manure and other materials that will charge the atmosphere 
with ammonia. Not only the use of ammonia but its possible as- 
sumption through the aerial structures suggest interesting studies. 
It is also interesting to speculate — since very few data are available — 
on the possible correlation between the natural habitat of a plant 
(for example, rice in a marsh) and the kind of nitrogen fertilizer 
to which it will respond. 

Organic Forms of Nitrogen as Plant Food. 

The stir that has pervaded soil-fertility circles for the last twelve 
years has brought forward the idea of the toxicity to plants of cer- 
tain organic soil constituents^ that all openminded students of the 
subject of fertility are bound to accept as a factor in the system. 
This activity has also been coincident with and in part has led to 
the establishment of the first important fact that is here emphasized, 
namely, that plants can use nitrogen in highly organized compounds, 
such as creatinine,* casein, and barbituric acid.^ 

This fact in itself is of large significance and should rearrange 
much of our teaching concerning practical soil management. It 
should give a new angle of vision on the use of stock and green 
manures. It should stimulate a close examination of the nitrogenous 

2 Kelley, W. P. The assimilation of nitrogen by rice. Hawaii Agr. Expt. 
Sta. Bui. 24. 191 1. 

3 Schreiner, Oswald, Reed, H. S., and Skinner, J. J. Certain organic con- 
stituents of soils in relation to soil fertility. U. S. Dept. Agr., Bur. Soils Bui. 
47. 1907. Further data on the subject are contained in later bulletins by Dr. 
Schreiner and his associates. 

4 Schreiner, Oswald, Shorey, Edmund C, SulHvan, M. X., and Skinner, J. J. 
A beneficial organic constituent of soils : creatinine. U. S. Dept. Agr., Bur. 
Soils Bui. 83. 1911. 

5 Hutchinson, H. B., and Miller, N. H. J. The direct assimilation of inor- 
ganic and organic forms of nitrogen by higher plants. In Centbl. Bakt., 2: 
513-547. 1911. 



FIITIN : LIVESTOCK AND SOIL ORGANIC MATTICK. 



99 



compounds in plants and in animal wastes with reference to their 
direct use by plants. It suggests a further reason for study of the 
micro and the macro plant organisms in the soil, viz., their effect on 
the form and solubility of the organic nitrog-enous substances. This 
is, of course, going- on in commendable fashion in several laboratories. 

Higher Plants Use Organized Carbonaceous Material. 

The second point to be emphasized is that the soil organic matter 
may contribute organized non-nitrogenous materials directly to the 
growing plant. It is now quite definitely established that plants can 
use many kinds of highly organized carbonaceous compounds pro- 
vided they are soluble, which solubility is of course essential to their 
transfer. The growing plant is able to use, to build into its new 
structure, ready-made molecules produced in an antecedent organism. 
It is like constructing a house of made-up parts, building up a book- 
case of sections. 

Knudson,^ of the Cornell Station, has reviewed the available litera- 
ture on this point and reports a considerable amount of investigation, 
showing that the common range of plants, such as corn, timothy, 
field peas, radishes, and hairy vetch can utilize such a wide range of 
materials as sugars, alcohols, aldehydes, and organic acids. The 
substances used in his investigation were saccharose, maltose, lactose, 
glucose, and galactose. With the possible exception of galactose, all 
these substances caused a marked increase in growth under sterile 
conditions. Hutchinson and Miller^ report relative to the assimila- 
tion of nitrogen compounds that "more or less satisfactory evidence 
of assimilation has been obtained," with a list of eighteen nitrogenous 
organic substances mentioned. 

Organic Matter a Source of Energy to Plants. 

An additional point in connection with the use of these organized 
materials is the energy relations of the system. Every organic mole- 
cule such as those mentioned represents the storage of a definite 
amount of energy. That energy is derived from light, commonly 
sunshine. If it is possible for a plant to use material elaborated by 
the light falling upon the leaves of a preceding plant, the new plant, 
using its own leaves to the full to receive the sunshine, should make a 
larger growth in a given time. 

* Knudson, L. Influence of certain carbohydrates on green plants. N. Y. 
(Cornell) Agr. Expt. Sta. Memoir 9. 1916. 
7 Hutchinson, H. B., and Miller, N. H. J. Loc. ext. 



lOO JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Knudson, in the paper previously mentioned, found that plants were 
able to make a measurable increase in dry matter under sterile condi- 
tions in the dark, thus indicating the direct use by the growing cells 
of the sugars studied. It is commonly recognized that the saprophytic 
organisms are dependent on this form of energy and that this rela- 
tion intimately controls such important processes as the fixation of 
nitrogen. That higher plants are also able to use such material is a 
most significant discovery. 

The practical aspect of this matter is the common custom of 
expert plant growers to load up their soil with freshly decayed 
organic matter. We never obtain the largest plant growth in a 
mineral soil, however large the supply of simple nutrients may be. 
The applications of manure in greenhouse and gardening practice 
bears no direct relation to the total soluble nitrogen and other nutri- 
ents supplied. The results suggest the sorting out and use of the 
suitable organized material that may represent only part of the 
soluble nutrients, not to mention the simple forms used under the 
influence of sunshine. 

I suggest the desirability of studying plant growth from the point 
of view of energy relations and the possibility of connection between 
this matter and the thermal sum for crop maturity.^ 

Effect of Animals on Organic Matter in Feed. 
We now come to consider the relation of livestock through their 
manure to the maintenance of soil organic matter. It is evident 
that if organic matter is important because of its organized structure 
and energy relations, as well as for its physical effects and supply of 
nutrients, a careful inquiry must be made into the effects of the 
animal on those organic materials. It has been customary to put 
the emphasis on the nitrogen, phosphorus, and potassium in the ma- 
terial, and by inference to suggest that the recovery of organic matter 
from the feed is similar in proportion to that of the nutrients. When 
the organic matter is considered, I suspect the average person is 
misled by the bulk of the material, made large by added water. 

Digestibility of Feeds. 

The newer method of rating the feeding value of materials in 
terms of energy units provides data from which this matter may be 
studied. Armsby and his associates have made determinations in 
the calorimeter at State College, Pa., of the distribution of the energy 

^ Bolls, W. L. Temperature and growth. In Annals of Botany, 22: 557- 
592. 1908. 



fiimmn: livi:st()( k and sou. oucanic mattku. 



lOI 



units in the feed consumed to the various body processes, and in the 
waste in both the undigested and the digested niaterial." These de- 
terminations cover a number of standard feeds and the use of several 
animals. Our interest in the results is the converse of the feeder's. 
We are concerned with the wasted residue. The energy or thermal 
value of the material very well suits our purposes of estimating the 



PERCENT 




T/MOTH 

CLO\/£f< 
HAY 

ALFALFA 

CORN 
STO\/£F^ 

WH£AT 
BRAN 

CORN 
MEAL 

HOAf/Ny 
CHOP 

GRA/N5 

M/X. HAY 
CORN M. 

ALFALFA 
GRA/U5 



Fig. 5. Graph showing minimum percentages of organic matter recovered in 
animal feces and urine, with the sums of the minima. 



organic matter available after an animal has consumed and digested 
its feed. Table i and figure 5 show the proportion of organic matter 
in the solid and liquid feces. 

There is an additional waste of energy in the bowel gases, mostly 
methane, amounting to 6 to 11 percent of the energy of the feed. 
Of this no account is taken, as it is assumed to be lost. 

From this table it is apparent that the animal returns only from a 
seventh to less than a half of the organic matter in the feed. The 
return is largest for the coarse, fibrous materials and lowest for the 

' Arm§by, H. P., and Fries, J. A. Net energy values of feeding stuffs for 
cattle. In U. S. Dept. Agr., Jour. Agr. Research, 3 : 435-491. 1915. 



I02 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

concentrates. The former are usually grown at home, the latter are 
largely purchased. This point bears on the importance of purchased 
feed in maintaining the soil. 



Table i. — Maximum and minimum percentages of organic matter in various 
feeds recovered in animal feces and urine, with the sums of the minima. 





Percentage recovered. 




Feed. 






Sum of minima. 




In feces. 


In urine. 






41.6-49.5 


3.9-4.6 


45-5 




41.0-42.7 


6.7-6.8 


47.7 




40.0-47.0 


5.5-6.4 


45-5 


Corn stover •. . . 


42.0-43.0 


4.0-4.5 


46.0 


Wheat bran 


30.0-33.0 


5.0-6.9 


35-0 




12. 0-15.0 


3.3-70 


15.3 




II. 5-13.0 


3.6-4.2 


15. 1 




18.0-27.3 


3.7-5.2 


21.7 


Mixed hay and corn meal 


20.0-28.0 


3.5-4.2 


23.5 




27.5-34.0 


4.3-5.3 


31.8 



Production of Manure. 

Comparison may be made of the digestion figures with the re- 
covery of organic matter in feeding practice. While there are many 
data on the production of manure, very few of the investigations have 
carried out all the weighings and determinations necessary to a 
conclusion. 

At Cornell University, R. E. Deuel investigated the production of 
manure and its several constituents by the college herd of 46 cows, 
The measurements covered a period of seven days and were the basis 
of a thesis for an advanced degree. The cows averaged 1,008 pounds 
in weight and were fed a ration that averaged 8.8 pounds of grain, 
30 pounds of silage, 10.2 pounds of low-grade alfalfa hay, and 4.2 
pounds of mangels. The grain consisted of cornmeal, distillers' 
grain, and bran, with a little cottonseed meal and oilmeal. The floors 
are covered with concrete and brick both in stable and lot. Shavings 
and sawdust were used for bedding and in addition the manure was 
dusted with land plaster in the drop. It was carefully weighed and 
removed from the stable twice a day and on three days was sampled. 
Both manure and feeds were analyzed. Following is a summary of 
the average manure production from each cow. 



Clear excrement produced daily 76 2 pounds 

Excrement produced daily, with bedding 83.4 pounds 

Yearly production of excrement per 1,000-pound animal 13.75 tons 

Organic matter consumed daily 21.3 pounds 

Organic matter voided daily 9.25 pounds 



FIIM»IN: LIVKSTOCK AND SOU. OKCANIC MATTKR. 



Organic matter regained 43.3 percent 

Nitrogen consnmed daily 0.59 pound 

Nitrogen voided daily 0.26 pound 

Nitrogen regained 44 3 percent 

Ash regained 63.6 percent 

Water in manure 81.8 percent 



The return of 43.3 percent of the organic matter can be approxi- 
mately checked by calculation frotii the digestibility tables of Armsby 
and Henry. The return of nitrogen, which appears low, is checked 
by calculation of the nitrogen in the milk produced. The return of 
mineral elements appears to be low and is not so well checked by the 
amount in the milk. 

Loss OF Organic Matter in Handling Manure. 

The organic matter in the feed must be followed further. There 
is a variable and unusually large loss in handling manure. Schutt^^ 
exposed mixed manure from cows and horses in bins, one of which 
was open to the weather, while the other was sheltered from the rain 
and snow. The loss of organic matter in six months was 65 and 50 
percent, respectively. Even when handled in the best way, there is 
a considerable loss which varies with the type of animal. Calcula- 
tions of the Ohio results^^ in the production of manure by 28 and 
30 steers fed on a clay and on a cement floor, respectively, for six 
months show a recovery of only 12.7 percent and 10.9 percent, re- 
spectively, of the organic matter in the feed consumed. 

The evidence is substantial that the average farmer does not return 
to the land more than 15 to 30 percent of the organic matter in the 
feed consumed by his animals. The loss from storage of such fer- 
mentive materials as horse manure is very large. Consider a heap 
of manure in the winter that- produces sufficient heat to melt frequent 
heavy snowfalls in the presence of temperatures around zero and that 
maintains an internal temperature hot to the touch. Contemplate the 
number of tons of coal required to supply this same amount of heat. 
The destruction of organic matter is as definite and effective in the 
heap as in the furnace for the same heat production. 

This is not necessarily an indictment against livestock husbandry. 
That business must rest on other factors than soil fertility. It sug- 
gests that perhaps the animal has been given too large a place in the 
maintenance of the soil. While the nutrients in manure are valu- 

Schutt, M. A. Barnyard manure. In Dept. Agr. (Canada) Central Expt. 
Farms Bui. 31. 1898. 
11 Thorne, C. E. Farm manures, p. 100. 1913. Orange Judd Co., New York. 



IP4 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

able if they can be saved, they are not indispensable. They can be 
purchased. The purchased feeds are usually not adequate to equalize 
the destruction of organic material by the animal. Concentrated 
feeds are most likely to be purchased. These are the materials 
on which the animal is most destructive. Therefore, while these 
purchased materials are an additional aid to keeping up the soil 
organic matter, they put off but do not necessarily avoid the evil 
day of a deficiency. 

The most critical factor in the system of soil management is organic 
matter. In view of the destructive action of animals, it may be 
better practice where the lack of organic matter is the controlling 
factor in yield to turn under a crop than to feed it. One crop in six 
may be turned under and the land be as well off as if the manure 
from those crops was applied. 

The question of the animal as a means of marketing crops must be 
considered. Where animal products are very high and crop products 
relatively low animal husbandry is most to be advised. On the other 
hand, where crops are relatively high in price and especially where 
the land is very poor, — if it is under cultivation — one can scarcely 
afford to keep stock ; but where the land is not cultivated but is 
merely grazed without any attempt to keep up its productiveness the 
situation is different. 

These conclusions are contrary in some respects to the usual teach- 
ings. It is a business proposition whether you will keep stock. 
The animals should be entirely able to pay their own way. When 
the manure must be closely figured to show a profit from keeping 
livestock, one is on doubtful ground. 

The purchase of feeds is an aid in the maintenance of the soil 
but if the loss is not made good where the feed is produced, it is a 
system of " borrowing from Peter to pay Paul " and is not good 
national economy. Animal husbandry must stand on its own bottom 
and pay its way to the soil, whether its supply of feed is produced 
on the farm where it is used or obtained from another part of the 
country. 

Summary. 

In conclusion, the points that should be emphasized are: 

1. The higher plants are able to use organized carbonaceous foods, 
both nitrogenous and non-nitrogenous. 

2. Carbonaceous food conserves energy in the process of growth 
of the crop and makes possible a larger total growth in a given time. 

3. The organic matter in the soil is the direct source of the car- 



MONTC.OMKRV : IIKATINC. SKKD ROOMS TO DKSTKOY INSKCTS. IO5 

bonaceous material used by the plant. Any process that permits the 
destruction of organic matter that might find its way into the soil is 
likely to be poor economy. 

4. Animals destroy from half to nine tenths of the organic matter 
in the feed consumed. It is burned up in the body processes and 
expended as energy. 

5. A further large loss occurs in the handling of the manure. 

6. It is entirely possible to maintain the organic matter in the 
soil without animal husbandry. On very poor soils, animal hus- 
band*ry may be bad practice. It may be justified by large profits from 
the animal products by means of which the loss of organic matter 
can be made up from other sources. 

Cornell University, 
Ithaca, N. Y. 



HEATING SEED ROOMS TO DESTROY INSECTS.^ 

E. G. Montgomery. 

The problem of protecting stored seeds and grains from insects is 
one that confronts almost every agronomist. Often hundreds of 
small lots are kept over from year to year or even for several years. 
If the seed room becomes thoroughly infested, it often requires con- 
stant care to prevent the total loss of these samples. 

Fumigation with carbon bisulfid or hydrocyanic acid gas is the 
most common method of protection used. This, however, has cer- 
tain disadvantages and is not always effective. Both of these gases 
are dangerous to human „ life and, therefore, are not always safe 
to use when the storage room is in a classroom or ofifice building. 
The fire danger from carbon bisulfid is so great that it cannot be 
used safely in large quantity, when it is likely to leak into adjoining 
laboratories or offices. The hydrocyanic acid gas is also corrosive 
and its continued use will generally destroy sacks and also the germi- 
nation of seed. 

Experience has also shown that fumigation is only partly effective, 
as it seldom kills all the eggs or larvae. The gas can not penetrate 
into closed containers such as cans or bottles, and penetrates only 
very slowly into large masses of grain. 

^ Contribution from the Department of Farm Crops, New York State College 
of Agriculture, Cornell University. Presented' at the ninth annual meeting of 
the American Society of Agronomy, Washington, D, C, November 14, 1916. 



I06 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

The heating method for killing insects has been well known for 
25 or 30 years, and in recent years has been advocated as a means 
of flour-mill fumigation. ^ The experimental work has shown that 
grain insects exposed to a temperature of 120° F. for 15 to 20 
minutes will be killed. This temperature also kills the eggs, larvae, 
and pupae. This complete destruction of all stages of the insects is 
a great advantage as one thorough treatment a year will ordinarily 
be suflicient. 

The Farm Crops Department at Cornell University fitted up a 
small seed store room with heating apparatus two years ago. We 
thought it would be a simple matter at first, but a number of problems 
developed. The room used was 8 feet by 18 feet by 10 feet high. 
The walls were light, temporary partitions, but were made tight with 
a coat of cement plaster. There are one window and one door in 
the room through which much heat escapes in cold weather, but there 
is little trouble in warm weather. 

For fumigating mills it is stated in both the Ohio and Kansas 
bulletins referred to that i square foot of radiating surface to 70 
cubic feet is sufficient to secure a killing temperature in midsummer. 
From 12 to 15 hours were needed, however, to attain this tempera- 
ture. In our store room, however, we have i square foot of radia- 
tion to 5.5 cubic feet of space and do not find it too much in winter, 
but it is more than is needed in summer. Probably a ratio of i to 10 
would be enough for the warmer months. 

The steam coils are connected to the regular steam radiator system 
used for heating the building. We also have it connected with a gas 
boiler for heating when steam is turned off in the summer time. 
However, we have found the gas boiler unnecessary as one or two 
heatings a year are suflicient. 

One of our greatest difficulties was to secure a uniform heat in 
all parts of the room. Thermometers were hung at different heights 
in the room and were also thrust into sacks and boxes of grain to 
ascertain how long it would take for all parts to reach a killing 
temperature of 120 to 130 degrees. We soon found that the tempera- 
ture near the ceiling was 20 to 30 degrees higher than near the floor. 
As the germinating quality of grain is destroyed at 150° F. and prob- 
ably injured at 5° less if long exposed, it was found necessary to 
introduce an electric fan to keep the air circulating. The fan alone, 

2 Goodwin, W. H. Flour mill fumigation. Ohio Agr. Expt. Sta. Bui. 234, 
1912. Dean, Geo. A. Mill and stored grain insect's. Kans. Agr. Expt. Sta. 
Bui. 189. 1913. 



MONTC.OMKRV: JIKATINC SKKI) KOOMS TO DESTROY INSECTS. I O/ 




I08 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

however, did not provide sufficient circulation until it was set near the 
ceiling, with, an 8-inch stove pipe at the back and extending to the 
floor (fig. 6). By this means the cooler air was sucked from the 
floor and delivered at the ceiling, resulting in a fairly uniform tem- 
perature in all parts. 

Further experiments showed that it took about lO to 12 hours for 
an outside temperature of 130° to raise the temperature in a 2-bushel 
sack of grain to 120° F. In closed boxes or cabinets a kiUing 
temperature would hardly be reached in 24 hours. To overcome this 
difficulty we made our larger grain receptacles (half -bushel size) 
with bottoms and tops of perforated metal. By placing these re- 
ceptacles on slatted shelves a fairly free circulation is secured, and a 
half bushel of grain will usually reach 120° in 5 or 6 hours with a 
room temperature of 130° F. 

It is of interest to note that mice are killed as well as the insects 
and that the germination of the grain has not been injured. Some of 
the ear corn on upper shelves was subjected to a temperature of 140° 
F. several times without apparent injury. After several years' ex- 
perience with chemical fumigation, I find the heating method very 
much more satisfactory and perfectly practical where it would be 
dangerous to use chemicals. 

Department of Farm Crops, 
Cornell University, 
Ithaca, N. Y. 



rii-:ri:us : t;Ri':i-:N manuring. 



109 



GREEN MANURING: A REVIEW OF THE AMERICAN 
EXPERIMENT STATION LITERATURE— 2. 

A. J. PlETERS. 

(Continued from the February issue.) 

Northeastern Division. 

The northeastern division, for the purposes of this paper, includes 
the states north of the Ohio River and east of the Missouri, except 
those bordering on the Atlantic Ocean. The Province of Ontario is 
also included in this division. Here, red clover is the chief legume 
used in the farm rotation, and is the only one entering into the experi- 
mental work of these stations to any great extent. 

CANADA. 

The most conclusive evidence we have regarding the value of red 
clover as a green manure is furnished by the work of the Central 
Experimental Farm at Ottawa. Although, so far as has been de- 
termined, the experiments usually ran for but one year, the number 
of tests made and the length of time during which these isolated 
experiments were carried on combine to make this a body of evi- 
dence of high value. Work on green-manure experiments is re- 
ported in various annual reports from 1893 to 1912 as well as in 
bulletins 40 and 165. 

For the purpose of this study the Canadian experimental farms 
may be considered in two groups. At the Central Experimental Farm 
at Ottawa and at the Nappan (N. S.) Farm the work with green 
manures and with legume catch crops has been almost wholly with 
red or mammoth clover. At the Western farms, — Brandon, Man. ; 
Indian Head, Sask. ; and Agassiz, B. C, — a number of legumes as 
well as some nonleguminous green-manure plants have been tried. 
At both the Central Experimental Farm and the Nappan Farm atten- 
tion was paid to the influence of the clover on the grain crop with 
which it was growing. (Reports 1895, 1896, 1897, 1898, 1903, and 
1904.) While the yields were often irregular, there was no evidence 
that the grain crop was especially affected by the clover. At the 
Agassiz Farm the yield of grain was less when clover was growing 
with it (Rpt. 1896, p. 440), but at both the Brandon and the Indian 



I 10 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Head farms the results were similar to those reported from the 
Central Farm. 

At the Central Farm in 1897 eight twentieth-acre plots were 
selected, four being seeded to grain with red clover and four to 
grain without the clover. The clover crop was turned under and 
oats sown in 1898. In 1899 barley was grown. Not only did the 
clover increase the yield of oats by nearly 30 percent on the average 
but the residual effect on the barley of 1899 was marked (Report, 

1899) . 

Other experiments gave similar results. In 1899 potatoes yielded 
28 percent better after clover than after carrots and in 1900 the 
yields of potatoes after a catch crop of clover with grain were uni- 
formly higher than after grain without the clover (Reports 1899 and 

1900) . In all these catch-crop experiments there was but one plot 
of a kind at a time, which naturally detracts from the value of the 
individual experiment. However, the fact that these trials were 
made for several years and in some of these years in several inde- 
pendent series of tests and that the results were uniformly in favor 
of the clover gives this evidence high value. 

It will be impossible to quote at length from the tables presented 
in these reports; a few will serve as examples. In 1902 oats, corn, 
and potatoes were grown on land on which grain, with and without 
a catch crop of clover, had been grown in 1901. The yields were 
much more after the clover than on the plots without clover, as shown 
in Table 4. (Report, 1902, p. 39.) 



Table 4. — Yields of oats, fodder corn, and potatoes on the Central Experi- 
mental Farm in 1902, on plots on which zvheat, barley and oats were grown the 
previous year, with and without clover. 



Previous crop. 


Oats. 


Fodder corn. 


Potatoes. 




Bushels 


Tons. 


Bushels 




63-53 


16.40 


353-33 


Wheat, clover 


72.94 


22.80 


396.00 


Barley, no clover 


61.18 


17-36 


346.67 




70.59 


23.60 


386.33 




58.88 


15.00 


358.33 




70.59 


20.40 


392.67 



The tenor of the reports for the following years is the same. In 
every instance the plot on which a catch crop of clover had been 
grown outyielded the other plot. At the Nappan Farm a similar line 
of work was carried on during the seasons of 1905, 1906, 1908, 1909, 
and 1 910. In practically every instance the yield was best after 
clover. 



PIETKKS : 



C.KKKN MANURING. 



The report for 1903, paj^es 33-37, gives the results of an extensive 
experiment consisting of several divisions, each having a number of 
plots, on some of which clover had been grown for one season (1900) 
and then turned under, on others for two seasons (1900 and 1901), 
while on check plots no clover had been grown in 1900. The yields 
of various crops were all larger during the subsequent three (and 
two) years on the clover than on the no-clover plots. Table 5 shows 
yields of corn, oats, potatoes, carrots and sugar beets in 1901, 1902, 
and 1903, following clover turned under. 



Table 5. — Yields of fodder corn in 1901, oats in 1902, and potatoes, carrots, 
and sugar beets in 1903 on the Central Experimental Farm on plots cropped 
to clover or without clover in preceding years. 



Previous crop. 


Fodder 

corn 
in 1901. 


Oats 
in 1902. 


1903. 


Potatoes. 


Carrots. 


Sugar beets. 


Clover in 1900 


Tons. 
25.80 
20.08 

27.22 
15.40 

27.88 
19.64 


Bushels. 
70.59 
58.82 
65.88 
70.59 
47.06 
72.94 
75-29 
51-77 
68.23 


Bushels. 
195-33 
175-33 
221.33 


Tons. 


Tons. 




22.30 
8.60 
"0. 


Clover in 1900 and 1901 

Clover in 1900 


31.48 
20.32 
21.30 






Clover, 1900 and i90i 





" No germination. 



In this and in other cases the effect of the clover was marked for 
three years. In 1904 records of yields for three years following the 
turning under of a clover catch crop in 1901 were available. Corn 
in 1902 was followed by potatoes and these by barley in 1904 and all 
yielded larger crops on the clover than on the no-clover plots (Report, 
1904, p. 34). 

The Nappan Farm (Report, 1905, p. 283) showed that a catch 
crop of clover in the grain turned under for grain the following year 
resulted in larger yields than where grain was grown continuously. 
The yields for 1905, the third year of the experiment, are given 
in Table 6. 



Table 6. — Yields of zvheat, oats, and barley in bushels per acre on the Nappan 
Farm in 1905, on plots continuously cropped to grain with and without clover 
for three years. 



Green manure. 


Wheat. 


Oats. 


Barley. 


Plot I. 


Plot 2. 


Plot I. 


Plot 2. 


Plot I. 


Plot 2. 


None 

Clover 


34.33 

40.00 


39-00 
41.67 


41.18 
55-29 


60.00 
60.85 


32.71 
37-92 


38.54 
42.92 



112 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Similar experiments with similar results are recorded in Report, 
1908, p. 261, and in Report, 191 1, p. 332. 

At the Central Farm records were also kept of the yields of grain 
on a series of plots that had been heavily fertilized with commercial 
fertilizers for about ten or eleven years. At the end of this time 
the fertilizers were no longer applied but a catch crop of clover was 
substituted. The records are found in the Report for 1903, p. 32. 
With heavy applications of complete fertilizers the average yield of 
oats for eleven years was 44.88 bushels. The average for the five 
years following, when no fertilizer was applied, but clover was turned 
under, was 56.69 bushels. The average annual yield of wheat was 
increased under this system by more than 40 percent and that of 
barley by 48 percent. While it is probable that in these cases, the 
clover served to make the residues of the previously applied fertilizers 
available, this is in itself a useful function. 

Besides the annual reports referred to, the subject of red clover 
as a fertilizer is discussed in Bulletin No. 40 of the Central Farm and 
by F. T. Shutt in Science for August 30, 1907, p. 265. The material 
in these publications is, however, identical with that of the Reports. 

Alfalfa and timothy sods were compared (Bulletin, Ontario Agr. 
College, No. 165) in 1900-1903. Sods of both were plowed each 
year and wheat, barley, and corn planted. The yields were as 
follows : 

1901 1902 1903 

Winter wheat. Barley. Corn. 

After alfalfa 61.5 bu. 30.2 bu. 24.0 bu. 

After timothy 42.1 bu. 19.7 bu. 17.9 bu. 

Although a discussion of the quantity of nitrogen added to soils 
by clover is not included in the scope of the present paper it may not 
be out of place to refer to one very instructive experiment recorded 
in the Canadian report for 191 1, p. 173. This experiment was 
started in 1902 at the Central Farm (see also F. T. Shutt's article in 
Science, Aug. 30, 1907, p. 265). A plot 16 feet by 4 feet was staked 
ofif and the sides protected by boards sunk to a depth of 8 inches. 
The surface soil was removed and new sandy soil the nitrogen 
content of which was 0.0437 percent was substituted. Superphos- 
phate and muriate of potash were added and red clover sown. Dur- 
ing each season the crop was cut twice and the material allowed to 
decay on the ground. Every second season the crop was turned 
under, the soil being stirred to a depth of 4 inches, and clover was 
again seeded the following spring. Table 7, taken from the report 
for 1911, p. 173, gives the data obtained during nine years of clover 
growth. 



PIETRRS : GREEN MANURING. 



Table 7. — Increase in nitrogen i)i soil due to groivth of clover, as shown by 
the percentage of nitrogen in water-free soil and the pounds per acre in the 
surface 4 inches. 



Period from beRinning of experiment. 


Date of col 


lect on. 


Percentage of 
nitrogen in water- 
free soil. 


Pounds of nitrogen 
per acre to deptli 
of 4 inches. 




May 13, 


1902 


0.0437 


533 


After two years 


May I \, 


1904 


.0580 


708 




May 15, 


1906 


.0608 


742 




May 30. 


1907 


.0689 


841 




May 23, 


IQ08 


.0744 


908 




May 4, 


1909 


.0750 


915 




May 5, 


I9II 


.0824 


1.005 








•0387 


472 



ILLINOIS. 

. At the Illinois station rotation plots have been carried on since 
before the organization of the station and are mentioned in 1888 in 
the first annual report as number 23 of the University of Illinois 
experiments. From time to time the results from these and from 
othfer plots added later have been reported or conclusions have been 
drawn from data based on the yields from these plots. 

In Bulletin 31 (p. 357), 1894, is reported the yield of corn and 
oats on ten plots for the years 1888 to 1893, inclusive. To some 
plots stable manure was applied and to others commercial fertilizers, 
while on some clover preceded the corn. The use of clover appears 
to have had nearly the same effect as the application of 24 loads of 
stable manure annually and the average yields of corn from the 
clover rotation plots was much better than from the corn-oats plots. 

In 1890 and 1891 there was corn on clover rotation plot 7, on 
corn-oats plot 4 and of course on the continuous corn plots i, 2, and 
3. Plot I received stable manure. The average yields for those 
years were: Plot i, 49.6 bushels; plot 2, 35.3 bushels; plot 3, 38.6 
bushels ; plot 4, 43.7 bushels ; and plot 7, 47.9 bushels. 

In 1888 and in 1893 there was corn on clover rotation plots 9 and 
10, on the corn-oats plot, and on plots i, 2, and 3. The average yields 
for the two years were as follows: Plot i, 45.3 bushels; plot 2, 38.3 
bushels ; plot 3, 37.9 bushels ; plot 4, 39.5 bushels ; plot 9, 51.6 bushels, 
and plot 10, 48.0 bushels. It is clear that the rotation including 
clover has maintained the fertility of the soil better than manure or 
com.mercial fertilizers and this effect is well marked in a poor corn 
year as 1893. The author states that the same was true in 1887. 

In Bulletin No. 42 (1896), p. 177, is given a table covering the 
yields on plots of Experiment No. 23, up to and including 1895. The 
increase in the corn crop when grown in a 3-year rotation was 20 



114 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



bushels per acre over the corn-oats rotation for the corn crops im- 
mediately following the clover and 15.2 bushels for the second year 
of corn after clover. Other references to Experiment No. 23 will 
be found in Bulletins 8, 13, and 37. 

In Bulletin 88 (1903) is reported work done on several experi- 
mental farms. Since the year 1903 was an unusually poor one for 
wheat the data are not of much value for our purpose except those 
from the Cutler field. It is shown (p. 140) that on this field the 
turning under of a crop of cowpeas in 1902 increased the net returns 
from the wheat crop of 1903 over that from a corresponding plot 
from which the cowpea crop was removed. The question whether 
the value of the cowpea hay turned under was more or less than 
the increased returns from the wheat is not discussed. 

Bulletin No. 99 (1905) contains the record of experiments on cer- 
tain fields on the Lower Illinois glaciation. On the Edgewood west 
field cowpeas were grown on one plot in 1896 and buckwheat on 
another and the crops were turned under for corn the following year. 
This was repeated in 1898 for the corn crop of 1899. The season 
of 1897 was so unfavorable that no conclusions can be drawn. In 
1899 on the undrained series of plots the corn crop on the cowpea 
plot of 1898 was more than 50 percent greater than on the sodium 
nitrate plot having the highest yield. On the drained series, however, 
one of the untreated plots returned the highest yield and the cowpea 
plot ranked fifth in a list of seven plots. In 1902 a catch crop of 
cowpeas was grown in the corn but the oat crop of 1903 showed 
no improvement. On the Odin field also the legume catch crop 
did not produce an increased yield of the following crops. It should 
be added, however, that the authors point out that the season of 
1902 was "very dry and the cowpeas made but little growth." 

On the Cutler field, reference to which was also made under 
the discussion of Bulletin No. 88, corn was grown in 1904 on the 
wheat plots of 1903. While the yield of corn was slightly better 
on the plot on which cowpeas were turned under in 1902 than on 
that from which the hay was removed, it was much less than the 
yield from the plot receiving manure instead of the cowpea vines. 
This was equally true of the plots with and without lime. It is 
presumed that 2 tons of manure per acre per year, all applied at one 
time on the corn, was used. The 1904 wheat crop was much injured 
by rust and consequently conclusions drawn from these plots would 
be open to objection. 

Further records from some of these experimental fields are given 
in Circular No. 97 (1905). The yields of wheat following a legume 



i»ii«:ti:ks: ckkkn manuring. 



115 



(cowpca) catch crop were sometimes less and sometimes more than 
those on the check plot, but the differences were mostly too small 
to be significant. In 1904 some plots on the Cutler field (page 9) 
received nitrogen (100 pounds per acre in 700 pounds dried blood) 
instead of the legume catch crop. A comparison of three plots re- 
ceiving a legume plus mineral fertilizers, including lime, with three 
in which nitrogen replaced the legume shows slightly higher yields 
for the nitrogen plots, but here again the differences are not marked 
enough to warrant any conclusions from a i-year test. 

On the Vienna field (p. 11) the turning under of a legume (cow- 
pea) crop appears to have increased the crop of wheat in 1905 very 
much, from 1.3 bushels per acre on the nonlegume to 10.8 bushels 
on the legume plots. These figures, with others, are also given in 
Soil Report No. 3 (1912), p. 9, and in Soil Report No. 11 (191 5), 
p. II. The increase in the yield of wheat is striking, especially so 
since the corn yield of both 1902 and 1903 was larger on the non- 
legume than on the legume plots, indicating if anything slightly 
better soil on the former. However, in 1909 cowpeas were again 
removed from Plot i and turned under on Plot 2, but the corn crop 
of 1910 was only 1.9 bushels higher on Plot 2 than on Plot i. In 
view of the fact that the results with wheat after turning under 
cowpeas on plots of series 200 and 300 were not so marked as on 
those of series 100 and in view of the small increase in the corn crop 
on all of these series after turning under a crop of cowpeas the large 
increase in the yield of wheat on plot 102 must be taken cautiously. 
The results do show a general increase of yield, however, on the 
plots on which the crop of cowpeas was turned under. 

Pot experiments with a worn soil were carried on from 1902 to 
1905 and are reported in Circular No. 97, pages 12 to 19. Part of 
these data are also given in Soil Reports No. 3, No. 7, No. 11, and 
No. 12. On some pots a catch crop of cowpeas was turned under 
with lime. The crop of wheat during 1903, 1904, and 1905 was 
markedly better on the legume-lime pots than on those receiving no 
treament, but how much of this gain is due to the lime can not be 
determined. One set of pots received a legume with lime, phos- 
phorus, and potassium while in another set nitrogen in commercial 
form was substituted for the legume. While the crops of 1903 and 
1904 were generally better and in some cases very much better from 
the "nitrogen" pots, that of 1905 was either better from the legume 
pots or as good as that from the nitrogen pots. In Soil Report No. 
12 the above data are repeated and the yields for 1906 and 1907 are 



Il6 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

included. These show that in general the legume continued to give 
better results than considerable quantities of commercial nitrogen. 

In Bulletin No. 115 (1907) further details are given concerning 
the experiments on the Vienna field. It is claimed that the use of 
legume green-manure crops (cowpeas) doubled the total yield of 
wheat for 1904, 1905, and 1906. The figures given are correct, but 
the increase in 1904 on the 200 series was negligible, and that of 1906 
on the 200 series was not large, while the greater part of this total 
increase was made on the 100 series, in 1905. No information is given 
as to the difference, if any, between plots of the 100 and 200 series 
but it may be questioned whether it is desirable to average figures 
taken from plots on different series. The yields of corn during 
1904, 1905, and 1906 were also slightly higher on the legume plots 
than on the nonlegume plots, though the total difference for the three 
years was slightly under 10 bushels. 

The pot-culture experiments mentioned above from Circular No. 97 
are again reported with the addition of the record for 1906. On 
the pots receiving phosphorus and potassium in 1906 the legume 
catch crop appears to have supplied the nitrogen, while where nitrogen 
and lime only was applied the yield was much lower than from the 
legume-lime pots. 

In Bulletin No. 125 (1908), Hopkins, Readhimer, and Eckhardt 
give an account of thirty years of crop rotations at the Illinois station. 
Part of the work reported is that on the oldest plots commenced in 
1879 and part on the new series laid out in 1895 and brought under 
definite systems of crop rotation and soil treatment in 1900. The 
record of the yields on the older series is also given by Hopkins 
in the latest edition of his book on Soil Fertility and Permanent 
Agriculture, p. 457. While working up these data the writer was 
able to secure also the records for 191 3 and 1914, so that tables 
presented below have been compiled from the records as given by 
Doctor Hopkins in his book, with the addition of those for the last 
two years. In 1904 the plots of the older series were divided and the 
south half planted to a legume catch crop each year together with 
lime, manure, and phosphorus. The north half had no treatment. 
Comparing the corn yields from the 2-year corn-oats rotation with 
those from the corn-oats-clover rotation it is evident that the yields 
have been maintained at a higher level by the latter rotation than by 
the former. The addition of the treatment to the south half of each 
plot also increased the yield but the presence of several other factors 
makes it impossible to draw conclusions in regard to the use of a 
legume catch crop on these plots. 



PIETEKS : GKEEN MANURING. 



117 



Table 8 gives the yields of corn on the plots of the 2-year and the 
3-ycar rotations, since 1904. While this record seems quite satis- 
factory so far as showing the benefit to corn, when the yields of 
oats for the same period are studied it is seen that the 2-year corn- 
oats rotation has given slightly better yields than the corn-oats-clover 
rotation. 



Table 8. — Yields of corn and oats in bushels per acre obtained in a 2-year 
corn-oats rotation and in a 3-year corn-oats-clover rotation, with and without 
a legume catch crop. 

CORN. 



Two-year rotation. 


Three-year rotation. 






Legume, 






Legume, 


Year 


Non-legume 


manure. 




Noneg ume 


manure. 


plots. 


lime, and 


Year. 


plots. 


lime, and 




phosphates. 




phosphates. 


19OS 


50.0 


44-9 


1904 


5.S.3 


72.7 


1907 


.47.8 


87.6 


1907 


80.5 


93-6 


1909 


33-0 


64.8 


19IO 


58.6 


83.3 




28.6 


46.3 


1913 


33 8 


47.8 


I913 


29.2 


25.0 








Average 


37-7 


53.7 




57-05 


74.35 


Average yearly increase for 3-year 












over the 2-year rotation 








19-35 


20 65 











OATS. 





34.7 


52.S 


1905 


42.3 


50.6 




32.9 


15-0 


1908 


40.0 


44-4 


I9I0 


33-8 


59.4 


1911 


20.5 


37-9 




55-0 


81.0 


1914 


39-6 


60.4 


I9I4 


33-6 


58.2 








Average 


38.0 


53-2 




35-6 


48.3 



On the new series of plots a legume catch crop (cowpeas at last 
working of corn or clover with oats) was introduced in 1901, but the 
record up to 1907 does not show any benefit from this treatment, for 
the " efifect has been a decrease as often as an increase " in the regular 
crop. 

For drawing conclusions as to the value of a legume in the rota- 
tion the data available in 1908 on the newer series can scarcely be 
used, for a corn crop was grown on all comparable plots in only one 
year. In that year (1904) the average yield of corn on all plots of 
the 3-year rotation (series 300) exceeded that on all plots of the 2-year 
series (500) by 29.5 bushels and the oats yield in 1895 on the 3-year 
rotation plots exceeded that in the 2-year plots by 7.8 bushels, a 
result different from that obtained with oats on the older series of 
plots. However, it should be noted that between 1898 and 1904 



I 18 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

the plots of series 300 produced but one corn crop while those of 
series 500 produced four. 

A comparison of the effect of a legume catch crop in oats with 
that of manure may be made for the years 1905 and 1907. In 1904 
and 1906 clover was sown in oats on the plots of series 400, numbers 
2, 4, 6, and 8. On number 4 lime was added, on 6 lime and phos- 
phorus, and on 8 lime, phosphorus, and potash. Plots 3, 5, 7, and 9 
are the corresponding plots where manure was used instead of the 
legume. The average yields of corn from the above plots in 1905 
and 1907 were as follows : In 1905 on all legume plots, 63.2 bushels, 
and on all nonlegume plots, 71.7 bushels; in 1907, 78.6 and 69.0 
bushels, respectively. The average yields for the two years were: 
On legume plots, 70.9 bushels, and on nonlegume plots, 70.2 bushels. 
Apparently the legume catch crop and the manure have been equally 
efficient. 

Circular No. 96 (1905) presents many of the same data given in 
Bulletin No. 125, and will not require special mention. 

Although not directly in line with this study it may be admissible 
to mention here an interesting experiment recorded in Bulletin 182 
(1915). In the course of a series of pot experiments with the 
"insoluble residue" from soil from which the acid-soluble potassium 
had been extracted, it was found that the clover failed or grew very 
poorly at first. By turning under what did grow and by improving 
the mechanical condition of the soil with sand from which the potas- 
sium had also been extracted, at the end of three years there was a 
fair growth and the total amount of potassium removed from the soil 
was many times greater than that which had been added in seed and 
impurities. The authors state that 99 percent of the potassium re- 
covered in the crop was secured from the " insoluble residue." This 
was made possible by two years of green manure. 

In Soil Report No. 14, Hopkins and others report on a test of 
green manures and commercial nitrogen on wheat. Wheat was grown 
in pots to some of which commercial nitrogen (amount not stated, 
but said to have been in excess of the value of the crops produced) 
was added. In other pots cowpeas were grown in late summer and 
fall and turned under for wheat. After the second year the crops 
raised after cowpeas were better on the average than those that had 
received commercial nitrates. 

INDIANA. 

This station appears to have conducted two long series of rota- 
tion and fertilizer experiments. The older, commenced in 1880, is 
mentioned in Bulletins 27 and 64. Besides continuous wheat culture, 



I'lKTKUS: CUKKN MANURING, 



119 



rotations of corn, oats, and wheat and of corn, oats, wheat, and 
clover-timothy were con(hicted. In Bulletin 64 (1897) the yields 
for the previous nine years are reported as having heen 6.22 hushels 
of corn more for the rotation with " grass " than for the grain rotation. 
At the end of the 9-year rotation period the entire field was put 
into corn. The corn in 1896 yielded an average of 5.66 bushels more 
on the "grass" rotation plots than on the grain rotation plots (Re- 
port, 1896). The gain in wheat is said to have been 7 bushels an acre 
(Annual Report, 1895).* 

In Bulletin 88 is a record of another experiment in which a corn- 
oats-wheat rotation was compared with a rotation of corn-oats-wheat- 
clover. Both rotations were fertilized at nearly, though not always 
at exactly, the same rate. The yields of corn, oats, and wheat for 
four complete rotations are recorded and the average annual yields 



are given below. 

Grain rotation. Grain and clover. 

Corn, 4 crops 50.52 bushels 34-34 bushels 

Oats, 4 crops 55-o8 bushels 45-62 bushels 

Wheat, 4 crops 30.99 bushels 20.37 bushels 



It is evident that the effect of the clover has been masked by that 
of the fertihzers, but it is not clear why the clover rotation should 
have given such poor returns. 

In Bulletin No. 114 (1906) what appears to be another rotation 
series is reported. Wheat was grown continuously or rotated with 
corn, with corn and oats, or with these and clover or with clover and 
timothy. The average wheat yields for the period of the experiment 
(18 years?) are as follows: 



Continuous wheat , 22.7 bushels 

Corn, wheat 16.5 bushels 

Corn, oats, wheat 22.0 bushels 

Corn, oats, wheat, timothy and clover 25.6 bushels 

Corn, oats, wheat, clover 29.1 bushels 

Corn, oats, wheat, clover 25.2 bushels 



In these experiments continuous wheat culture appears to have 
given better results than a grain rotation. The introduction of clover 
into the rotation caused some increase in the yield of wheat. The 
average of the last three wheat yields was 16.9 bushels on the corn- 
wheat and 22.1 bushels on the corn-oats-wheat-clover rotation. 

* References to this series of experiments of more or less importance are 
found in the following bulletins and reports: Bulletins 16, 1888; 27, 1889; 32, 
1890; 39, 1892; 43 and 45, 1893; 50 and 51, 1894; 55, 1895; 64, 1897; Annual 
reports for 1888, 1894, 1895, and 1896. 



I20 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

IOWA. 

Brown in Research Bulletin 6 (1912) presents a study of the effect 
of rotations on the bacterial activity of soils. This work was carried 
on in a field that had been in various rotations from 1907 to 1911 and 
the yields for 191 1 from the various rotations are given. It appears 
that the plots having a corn-oats rotation with a cowpea catch crop 
returned the lowest yield, a trifle less than the continuous corn plot. 
The corn-oats-clover with clover catch crop returned 20 bushels per 
acre more than the lowest plot. The results have little value, how- 
ever, as the time covered (4 years) is short for the effect of rota- 
tions to become evident. 

The yields of corn for 1912 and 1913 are given in Research Bul- 
letin 25 and these show that in 191 2 the best yield was obtained from 
the plots on which a clover catch crop was turned under and that a 
cowpea catch crop was followed by yields nearly as large as those 
taken from the 3-year rotation plots. In 1913, however, the 3-year 
corn-oats-clover rotation gave the largest yields, followed by the plot 
on which rye had been turned under. The yields from the clover 
and the cowpea catch crop plots were less than where rye had been 
turned under. 

In Bulletin 161 it is stated that for Iowa soils yields are better in 
rotations than in continuous culture, but it was also found that cow- 
peas sown at the last working of the corn depressed the yields o-f corn 
slfghtly over the 8-year period. During the first 4-year period the 
yield after a cowpea catch crop was slightly better than from the 
check plots, but during the second 4-year period the reverse was the 
case. When manure or manure and cowpeas were used the cowpeas 
depressed the yield during both periods, but when phosphorus was 
added slightly higher yields were obtained from the cowpea plots than 
from those with manure and phosphorus but without cowpeas. 

Oats yielded slightly better for the 8-year period on land having 
had a cowpea catch crop than on the check, but when manure or 
manure and phosphorus were added to the cowpeas the yields were 
lower on the cowpea plots than on those without cowpeas. The 
authors attribute this depressing effect to " unsatisfactory moisture 
conditions and lack of available plant food." 

The effect of turning under various catch crops and of the use 
of clover as a crop in a 3-year rotation of corn-oats-clover is shown 
in Bulletin 167 (1916). Besides the 3-year rotation there were four 
2-year rotations of corn and oats, one without a catch crop, one in 
which clover was sown with the oats, and two in which cowpeas and 



IMKTKRS : r.RKEN MANURING. 



121 



rye respectively followed the oats as catch crops. All three catch 
crops were turned under in the fall preceding the corn. It is stated 
that satisfactory quantities of green material were plowed under in 
each case. Comparing the yields on the 2-year rotations it is noted 
that the use of cowpcas and of rye did not increase the corn yields 
during the six years of the experiment, but that the yield following 
the clover catch crop was more than ii bushels better than that on the 
plain 2-year rotation. The yield of oats was practically the same on 
the 2-year rotations with catch crop and on all was slightly better 
than that from the check rotation. The yield of corn from the 3- 
year rotation was a little less than that from the 2-year clover catch 
crop plots, but the yield of oats was more than 10 bushels better than 
that from the 2-year catch crop rotation. The financial results given 
show that the 3-year rotation was most profitable, closely followed by 
the 2-year rotation in which a clover catch crop was used. 

MICHIGAN. 

In 1896 this station started an experiment which we must regret 
has not yielded more positive results. The plan as given in the Report 
for 191 1, p. 212, was simple but calculated to answer fundamental 
questions. On one series wheat was grown continuously, on two 
plots wheat alternated with clover, on two orchard grass was grown 
continuously, and on two other plots clover was to have been grown 
continuously. There were also two fallow plots on which no crop 
was grown for ten years. Some other rotations were also tried, as 
well as continuous corn culture. There are indications that the land 
was not uniform or if so that the method of treatment left much to be 
desired. From 1890 to 1896 all plots were cropped the same so as 
to determine the uniformity of the soil. But in 1896 wheat yielded 
2,620 pounds, 3,440 pounds, and 2,300 pounds on plots 62, 64 and 
66, while the average yield of plots i, 7, 36, 60, and 68, was but 11 1 
pounds. No explanation is offered for these wide variations, but 
the work was carried out on this land. After ten years of cropping 
as above stated all plots were planted to corn in 1906, oats in 1907, 
and wheat in 1908, to determine the efifect of the previous cropping 
on fertility. During these 10 years the clover had failed twice on 
the wheat-clover rotations and there is no record of a clover crop for 
six years out of ten on plot 34 and for four years out of ten on plot 
72, both continuous clover plots. Whether the clover failed (and if 
so, how the plot was treated) or whether the record was lost is not 
stated. The average yields in pounds per acre of all plots of a kind 
in 1906, 1907, and 1908 are given in Table 9. 



122 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Table 9. — Yields of corn in igo6, oats in 190/, and wheat in igo8 after various 
rotations at the Indiana station. 



Plot No. 


Previous crop. 


Corn. 


Oats. 


Wheat. 


I, 66 


Wheat, clover 


5.000 


. 950 


1. 515 


7, 36, 60, 68 


Wheat continuously 


3.938 


568 


1,008 


17. 19 


Beans continuously with rye turned 










under each year 


5.055 


825 


925 


21, 27 


Corn continuously 


3.040 


990 


1.665 


23. 25 


Orchard grass continuously 


5.900 


1,160 


2,095 


34. 72 


Clover continuously 


4.320 


755 


1,610 


29. 31 


Fallow continuously 


5.080 


950 


1,680 



The imperfection of the record, especially for the continuous clover 
plots, makes an interpretation of the figures of doubtful value, but 
in general the fact seems to stand out that the wheat-cover rotation 
has maintained the fertility of the land better than continuous wheat 
culture and perhaps a little better than continuous bean culture with 
rye as a green manure. If there was clover every year on plots 34 
and 72 it would seem as if orchard grass was better than clover, but 
the lack of records makes it impossible to draw conclusions in regard 
to this. 

MINNESOTA. 

At this station four series of rotations were established in 1894. 
These are outlined in Bulletin 40 (1895), referred to in the annual 
reports for 1895 1896, and are more fully discussed in Bulletin 
109 (1908). The review below refers to the latter bulletin. The 
plots in Series I cover 5-year rotations in which corn, wheat, oats 
follow in this order with two years of grass or clover between the 
wheat and oats. The rotations differ only in respect to the meadow 
crop used. Plots i, 6, and 11 had clover and timothy; plot 2, brome 
grass ; plot 3, timothy ; and plot 4, clover. All plots received 8 tons 
of manure to the acre on the corn. The rotation for plot 4 is given 
on page 303 as a 4-year one, and on page 309 as a 5-year one. It 
is understood that the rotation was a 4-year one. 

In Series III wheat is the indicator, being grown continuously in 
plot 2, and continuously with 6 pounds of red clover seed sown per 
acre as a catch crop in plot 3 ; on plots 4 and 5 wheat was rotated 
with clover, the second crop of clover being plowed under on plot 4 
and removed from plot 5. No manure was used on any plot of 
Series III. The other plots of this series and Series II and IV are 
of no value for our present study. Unfortunately the value of the 
results from Series I is practically destroyed because the proposed 
sequence of crops was not adhered to. The average yield of wheat 



riKTi:us: (;k1':kn manukinc. 



123 



was higher from plot 4 (clover) than from any other plot in this 
series, exceeding that from the next highest plot (timothy and clover) 
by •3.4 bushels, while the yields from the brome grass and from the 
timothy plots were between 4 and 5 bushels lower than those from 
the clover plot. It must be noted, however, that because the pro- 
posed course of rotations was not adhered to, one year of wheat 
preceded meadow and that following this one year of meadow there 
were seven grain years, three of which were wheat. It would seem, 
therefore, that the better yield from plot 4 must represent a better 
soil condition, or be due to some other cause than the roots and 
stubble of the one year of clover meadow. Plot 3, in which timothy 
only was used as a meadow crop, yielded nearly as much wheat as 
the average of plots i, 6, and 11 (timothy and clover). This plot 
also yielded 4.3 bushels of oats and 7.7 bushels of corn more than 
the timothy and clover plots. 

In Series III, plots 2 and 3 had wheat every year, but on plot 3 
clover was sown with the wheat and turned under as a catch crop. 
The average annual yield from plot 3 exceeded that from plot 2 by 
nearly 3 bushels. In a letter dated January 12, 1916, Professor Boss 
informs the writer that the Minnesota station is still carrying on 
these two continuous wheat plots and that the addition of 6 pounds 
of red clover seed to the wheat has given us an increased yield of 
wheat of about 2.6 bushels per acre over the entire period." Plots 
4 and 5 were treated alike save that on plot 4 the second crop of 
clover was plowed under and this was removed from plot 5. In 
spite of this greater amount of green matter turned under the average 
annual yield of wheat was 2.3 bushels less on plot 4. It would be 
expected that the greater amount of green matter turned under on 
plot 4 would have resulted in increased yields. No explanation is 
forthcoming for the condition recorded. 

A large part of this bulletin is devoted to showing that clover in 
a rotation tends to maintain the nitrogen content of the soil. While 
the record is not as harmonious as could be wished the evidence for 
the above mentioned view seems sufficient. The nitrogen content as 
well as the average annual yields for plots 2 and 3 of Series III are 
as follows: 



Yield, bu. 



Nitrogen present. 



Plot 2, wheat continuously, no clover 
Plot 3, wheat, clover catch crop 



18.57 
21.8 




Here not only has plot 3 produced more every year but the nitrogen 
content has gradually increased while that of plot 2 has decreased. 
Another series of rotations is recorded in Bulletin 89, but it is not 



I 24 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

possible here to separate the influence of the manure added from that 
of the clover. The rotation plots outyielded those continuously in 
wheat but this would probably have been the case had there been 
no clover in the rotation. Other bulletins, as Nos. 53 and 128, treat 
of the gain and loss of nitrogen and humus under various conditions. 

NEW YORK (CORNELL), 

A comparison of the yield of corn and of oats for one year on 
land that had been in alfalfa for six years with that on a contiguous 
plot previously in timothy for six years is made in Bulletin 339. On 
the alfalfa sod corn in 1912 yielded 62 bushels per acre; on the 
timothy sod, 47 bushels. In 191 3 oats were planted, the yield being 
26 bushels on the alfalfa land and 27 bushels on the timothy. The 
growth of corn is said to have been notably greater on the alfalfa 
than on the timothy sod. 

OHIO. 

With the exception of an observation on sweet clover (Bui. 42) 
and wire-basket experiments described in Bulletin 168, no green 
manure work has been reported from this station. There is, how- 
ever, a great mass of literature on rotations and the effect of fertilizers 
and from the data presented certain conclusions may be drawn. It 
must be insisted, however, that the evidence on this point is all in- 
direct, the rotations being such that direct comparisons are not 
possible. 

Wheat grown on sweet clover sod in 1891 yielded 26.9 bushels per 
acre while on adjoining land that had been in corn, oats, or wheat 
during the preceding four years, the yield was 18.6 bushels (Bui. 42). 

The value of having vegetable matter in the soil is brought out in 
Bulletin 49, where the results from adding phosphates were much 
better on land rich in organic matter than on that poor in organic 
matter. 

Various rotations covering five or three years have been carried on 
at the Wooster station and also at some of the substations, and com- 
parisons made with continuous culture. Wheat on both fertilized 
and unfertilized plots averaged 5.6 bushels per acre better in three 
years on the rotation plots than on the continuous culture plots (Bui. 
53). Corn yields on fertilized plots in continuous culture declined 
30 percent in 20 years, while on the fertilized plots in the rotation 
series the gain in yield was 23 percent (Bui. 282). A 3-year rota- 
tion of corn-wheat-clover produced better results than a 5-year one of 
corn-oats- wheat-clover-timothy (Circ. 120). Wheat, oats, and corn 
were all grown in continuous culture for 20 years and in 3-year and 



PIETKKS: CRICKN MANURING. 



£25 



5-year rotations for 17 and 20 years respectively, in which the corn 
followed clover. All three crops yielded best in the rotation series, 
corn and wheat giving largest yields in the 3-year rotation, the yield 
of corn being more than twice as large from the 3-year rotation as 
in continuous culture (Circ. 144). Tobacco grown on clover sod in a 
3-year rotation of wheat-clover-tobacco yielded more than 50 percent 
more than on unfertilized land continuously cropped to tobacco 
(Bui. 285). It is not possible, however, in any of the above work 
to separate the benefit, if any, from legumes from that which would 
accrue merely from the practice of a rotation. 

That clover sod may furnish the nitrogen needed for at least the 
next succeeding crop is shown in Bulletin 182 and in Circulars 40 and 
79. In a rotation of potatoes-wheat-clover, some plots received 
potash and phosphoric acid but no nitrogen while others received 
varying amounts of nitrogen in addition to potash and phosphoric 
acid. All plots were on clover sod, and the largest yields were on the 
plots not receiving nitrogen. In one series (Circ. 40) the yields of 
wheat following the potatoes were also fully as good on the plots 
with no extra nitrogen as on those with extra nitrogen. In other 
cases (Circ. 79) the nitrogen left by the clover appears to have been 
insufficient for both potato and wheat crops. Soybeans are said to 
have increased the yield of wheat (Bui. 275, p. 310). 

In the wire-basket experiment (Bui. 168), 5 percent of chopped 
clover was added to the soil, either alone or with lime. Three suc- 
cessive crops of wheat were grown and the rate of transpiration 
recorded. The green weights were obtained for the first two crops. 
The first crop was planted 10 days after the clover was added to 
the soil and both the transpiration rate and the green weights were 
depressed by the addition of the green manure. A second crop was 
then grown on the same soil without disturbing it and this showed 
decided benefit from the green manure, this benefit amounting to 
about 30 percent. In another wire-basket experiment a similar result 
was obtained. The first crop was depressed as a result of adding 
the green manure while the second crop was benefited. 

WISCONSIN. 

Potatoes were found to yield more on plots on which a heavy 
stand of red clover had been turned under than on plots receiving 
barnyard manure or commercial fertilizers (Bui. 147). The yields 
of marketable potatoes in bushels per acre were as follows : Barnyard 
manure, 183.5 bushels; commercial fertilizer, 183 bushels; no fertil- 
izer, 168 bushels ; and clover turned under, 234.5 bushels. 



126 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

SUMMARY. 

Wheat and corn have been the chief indicator crops in the north- 
eastern division, with oats and potatoes occupying subordinate posi- 
tions. With a few exceptions red clover has been used as the legume 
crop for green manure and for rotation. In Illinois, cowpeas have 
been used occasionally and there are a few records of trials on alfalfa 
and on timothy sod. 

In most of the Canadian experiments red clover has been used as 
an annual green manure catch crop, being seeded with wheat and 
turned under for the next year's wheat. As such, as well as in all 
the other experiments, red clover proved beneficial to the succeeding 
crop. In the Canadian experiments, red clover was also shown to 
have a marked residual value. Even when used as an annual catch 
crop the effect of turning under clover was frequently noticeable for 
three successive crops. 

At the State stations in this section, there have been few real 
green manure tests. Minnesota found that a catch crop of clover in 
wheat was a benefit. Illinois found no benefit from a cowpea catch 
crop in corn, and in some cases none when cowpeas were grown as 
a regular summer green manure crop. Most of the work with red 
clover has been in rotation experiments, and it is difficult to interpret 
the results with any great degree of clearness. Rotations having 
clover as one member yielded better than similar rotations without 
clover, or than continuous cropping, but a part of this result was 
doubtless due to the better suppression of pests. Even making some 
allowance for this, however, clover may safely be credited with a 
beneficial effect upon the yield of corn and wheat in rotation. 

That clover and cowpeas were able to furnish the nitrogen for a 
succeeding crop was shown by both Ohio and Illinois in both pot 
and field experiments, but the marked residual value of the clover 
found in Canada was not evident in Ohio. 

The value of much of the State work in this section for determin- 
ing the value of clover is seriously impaired by reason of changed 
plans or complications in the rotations that make it unsafe to draw 
positive conclusions. 

As a whole it can be said that the results obtained at the State 
stations in the Ohio and Mississippi valleys have been less striking 
than those reported from Canada and have often been negative or 
the evidence has been of doubtful value. 



{To he concluded in the April issue.) 



ARNY AND TIIATC IIKR: INOCULATION OK AIJ-ALKA. 127 



THE EFFECT OF DIFFERENT METHODS OF INOCULATION 
ON THE YIELD AND PROTEIN CONTENT OF ALFALFA 
AND SWEET CLOVER— 2.^ 

A. C. ArNY AND R. W. TlIATCHER.2 

In a previous article,^ we reported the results of the first year's 
work (on the crop of 1914) on a study of the effect of different 
methods of inoculation at seeding time upon the yield and composi- 
tion of alfalfa and sweet clover grown in subsequent years on the 
treated and adjacent untreated check plots. We have now completed 
a second year's work, on the crop of 191 5. The results of the two 
seasons' work are, in the main, so concordant and the conclusions to 
be drawn so plain that we desire to present now the second set of 
data, together with our conclusions concerning the problems which 
have been under investigation. 

Effect of Different Methods of Inoculation on the Yield and Composition 

OF Alfalfa. 

First Series, Fields E and F ; Commercial Culture versus Inocula- 
tion with Soil, with and without Liming. — Full descriptions of the 
soil conditions, size of plots, methods of seeding, and methods of 
inoculation used in the series were given in our first paper and need 
not be repeated here. The only variation from the method of pro- 
cedure described in that paper was that in 191 5 the samples for 
determination of dry matter and nitrogen content were taken from 
the green material as it was cut in the field instead of from the air- 
dried hay. The weights from which the yields of dry matter per 
acre are calculated were, therefore, the weights of green material 
as cut instead of those of cured hay. 

The plots of Field E were seeded during the spring and summer 

1 Contribution from the Minnesota Agricultural Experiment Station. Pre- 
sented by the junior author at the ninth annual meeting of the American Society 
of Agronomy, Washington, D. C, November 14, 1916. 

2 The thanks of the authors are due to Messrs. R. A. Thuma and Shinjiro 
Sato for most of the analytical work from which the data presented in this 
paper were prepared. 

3 Arny, A. C, and Thatcher, R. W. The effect of different methods of inocu- 
lation on the yield and protein content of alfalfa and sweet clover. JouR. 
Amer. Soc. Agron., 7: 172-185. 1915. 



128 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

of 1912. The crop of 191 5, therefore, represents the fourth season*s 
growth, or third harvested crop from these plots. The three cuttings 
of this crop were made on June 24, July 22, and September 25, 
respectively. The plots in Field F were seeded in the spring of 191 3 
and the crop here dealt with is the third season's growth, or second 
crop from these plots, the three cuttings being made on June 22, 
July 29-30, and September 20. 

The total yields of dry matter per acre, the average protein content 
in the dry matl^r from each of the three cuttings, and the total yield 
of protein per acre from each of the three plots which received the 
same treatment at seeding time and the average of these yields for 
each kind of treatment are shown in Table i. 



Table i. — Effect of different methods of inoculation at seeding time {1912 and 
1913) on yield and protein content of alfalfa grown in 1915. 
(Field E, Series I, Plots 11-16; Field F, Series III, Plots 1&-62,) 



Kind of inoculation. 


Plot No. - 


Dry 
matter 
per acre. 


Average 
protein in 
dry matter. 


Protein 
per acre. 






Pounds. 


Percent. 


Pounds. 




II 


9.632 


22.34 


2,152 




18 


8.487 


22.42 


1.905 




52 


8,529 


20.93 


1.785 




Average 


8,883 


21.93 


1.947 




12 


9.343 


22.90 


2,140 




24 


8,656 


22.52 


1.949 




58 


8,487 


21.95 


1,863 




Average 


8,829 


22.46 


1,984 




13 


9,110 


23.38 


2,130 




25 


8,561 


23.68 


2,027 




59 


8,171 


22.22 


1,816 




Average 


8,614 


23.10 


1,991 




14 


9.426 


22.13 


2,085 




26 


8,226 


22.71 


1,868 




60 


8,205 


21.18 


1.738 




Aveiage 


8,620 


22.01 


1,897 


Soil from alfalfa field +2 tons limestone per 












15 


9,908 


22.16 


2,197 




27 


8,218 


23.45 


1,928 




61 


8,272 


22.61 


1,870 




Average 


8,799 


22.74 


1,998 


None, +2 tons limestone per acre . 


16 


9,402 


20.8-3 


1,964 




28 


8,028 


22.11 


1.775 




62 


8,631 


21.12 


1,823 




Average 


8,687 


21.33 


1,854 



The results of the preceding year* showed that on the plots in 
Field E, on a well-manured soil, there was no uniform effect of 
inoculation, either with or without liming, upon the yield or protein 
content of the crop in the third year after seeding. In Field F, how- 

* Loc. cit., Tables i, 2, and 6. 



ARNY AND TllATC HKK: INOCULATION OK ALFALFA. I 29 

ever, on a soil of only medium productivity, in the second year after 
seeding there was a very considerable increase in the yield of dry 
matter and a significant increase in the percentage of protein (the 
combination of these two factors resulting in an increased yield of 
protein per acre) on those plots which had been inoculated with soil 
from an old alfalfa field and limed. Both the inoculation alone and 
the liming alone produced significant increases in growth, while the 
combination of inoculation and liming produced the largest yields 
of all. This year's results, on the other hand, do not show any in- 
creased yields of either dry matter or protein on the inoculated plots 
in either of the fields. The variations between plots are not greater 
than would be expected in duplicate plots in the same field which 
receive the same treatment. 

There are three possible explanations for - the coming of these 
plots to a uniform yielding capacity in the third year after seeding. 
First, the inoculation may have spread from the plots which were 
inoculated at seeding time to the untreated check plots in the same 
series. Second, the heavier draft of the rank-growing inoculated 
crop upon the plant food in the soiP may have reduced the productive 
capacity of the soil on the inoculated plots to a par with that of the 
uninoculated plots. And third, the plants on the uninoculated plots 
may have become so thoroughly well rooted and established by the 
third season that they are able, without the aid of the bacteria, to 
make as luxuriant a growth as are those on the inoculated plots. 

A comparison of the data presented in Table i with the same data 
for the crop of 1914 from these same plots indicates very clearly 
that the first of these explanations is the correct one. The yields of 
dry matter from all the plots in 191 5 were uniformly high, which 
contradicts the idea that the yields on the inoculated plots have come 
down to the basis of lack of inoculation. Furthermore, it is clear 
that the percentage of nitrogen (protein) in the crop from the unin- 
oculated plots has increased up to the average of that for the in- 
oculated plots, which in itself is a good indication that the plants on 
the plots which were not inoculated at seeding time are now as wel\ 
supplied with the nitrogen-fixing bacteria as are those on the inoculated 
plots. 

A search of the literature dealing with alfalfa inoculation experi- 
ments has failed to give any other evidence bearing upon this problem. 
Hutton^ reports yields of alfalfa in 1910 at the Lacombe Farm, on 

5 Loc. cif., Table 11. 

^ Hutton, G. H. Comparing inoculated soil with culture as a means of inocu- 
lating for alfalfa. Canada Expt. Farm Rpt. 191 1, p. 496. 



130 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

plots seeded in the spring of 1909 with different methods of inocula- 
tion, but no yields in subsequent years, on these same plots, have been 
reported. Other experimenters report " success " or " improvement 
in obtaining a stand of alfalfa during the first season after the ex- 
perimental inoculation. 

Second Series, Quinn Farm, Inoculation of Alfalfa by Transfer 
of Soil from Sweet Clover and Alfalfa Fields. — These plots were 
seeded in April, 191 3, on soil which had never grown alfalfa, nor 
had this crop ever been grown within half a mile of the field. The 
crop of 1914 showed a very large gain in yield of dry matter per 
acre^ and in percentage of protein in the dry matter^ from, those ploits 
which had been inoculated at seeding time, soil from an alfalfa field 
and from a sweet clover field being equally efficient in producing both 
these results. 

The yields of dry matter and protein content of the crops from 
these same plots in 191 5, from the three cuttings made on June 16- 
17, August 7, and September 24, respectively, are shown in Table 2. 

Table 2. — Effect of inoculation of alfalfa when sown in IQ14 with soil from 
sweet clover field and from alfalfa field on the yield and protein content of the 
crop in 191 5. 

(Quinn Farm.) 



Plot 
No. 


Kind of inoculation. 


Dry matter per 
acre. 


Protein in dry 
matter. 


Protein per 
acre. 






Pounds. 


Percent. 


Pounds. 


12 


None 


3.129 


19.87 


1. 136 


13 


With sweet clover soil 


3.824 


20.97 


1.366 


14 


None 


3.615 


19.84 


1.295 


17 


With alfalfa soil 


4.423 


21.61 


1,604 


20 




3.301 


19-45 


1,320 



The results presented in Table 2 clearly show that the effect of 
inoculation at seeding time (in 191 3) is much less in the crop of 
1 91 5 than in that of 1914. There is, however, a significant increase 
in yield of dry matter and in the percentage of protein in the dry 
matter on the inoculated plots, as compared with the adjacent unin- 
oculated ones, even in this third year after seeding. It appears from 
these results, as well as from those of the preceding year on plots 
which had been inoculated two and three years previously, that the 
effect of inoculation at seeding time upon the chemical composition 
of the resultant crops is very marked in the first crop (second season's 
growth), comparatively slight but still noticeable in the second crop 

7 Arny, A. C, and Thatcher, R. W. Loc. cit., p. 177, Table 3. 
^Loc. cit., p. 178, Table 4. 



ARNY AND TIIA1\:HKK: INOCULATION OF ALI'AhFA. I3I 

(third season's growth) and disa])i>cars entirely in suhscciucnt crops. 
This is undoubtedly due to the spread of the inoculating- bacteria from 
the inoculated plots to adjacent uninoculated ones. It ap|>ears that, 
under the conditions at University Farm, this takes place to a very 
considerable extent during the second season's growth and that by 
the close of the third season the uninoculated plots in any given series 
are sufficiently supplied with bacteria so that subsequent crops are 
uniform in yield and protein content, at least so far as inoculation 
is concerned. There seems to be no need, therefore, to carry any 
studies of the effect of inoculation beyond the second season's growth, 
under our conditions. On the other hand, if further studies of this 
kind are projected, they irmst provide for comparison plots located 
at such distances apart that there can be no possibility of transfer of 
inoculation from one plot to another. 

The difference in yield in 191 5 between the plots inoculated with 
alfalfa soil and with sweet dover soil is probably due to other causes 
than the difference in character of the inoculating material, as results 
in other fields clearly contradict the assumption that inoculation with 
alfalfa bacteria is more effective liian that with sweet clover bacteria. 

Effect of Inoculation on Yield and Composftion of Tops and Roots of 
Sweet Clover and Alfalfa. 

The results of the first year's study of this problem gave some 
very conclusive evidence concerning the effect of inoculation upon 
the total yield of dry matter per plant and upon the ratio of tops to 
roots in both alfalfa and sweet clover plants, but left some uncer- 
tainty as to the effect of the inoculation upon the chemical composi- 
tion of the different parts of the plants, especially with regard to 
certain of the mineral constituents in the root^. It was thought that 
this might be due in part to the difficulty which was experienced in 
getting the fine fibrous roots wholly free from soil particles, and in 
part to the fact that the small amount of material which was avail- 
able did not permit as large a number of analytical determinations 
as was desired. On this account, we decided to repeat the experi- 
ments. The desired inoculated and uninoculated check material was 
available and, with the experience of the preceding year as a guide, 
much more satisfactory sampling of the plant growths was obtained. 
This year's results present very convincing evidence on all the points 
in question and the data are presented with the conclusions which 
we have drawn from them. 

The method of measuring the sample square-yard areas and of 
securing the entire growth of tops and roots was essentially the same 



132 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



as that described in our former paper.^ With added experience, 
however, it was found possible to collect more completely the fibrous 
root growth of the plants and to free this almost entirely from soil 
particles, so that the samples as received at the laboratory were very 
good representatives of the crops as they grew in the field. 

In the field from which the sample square yards of sweet clover 
were taken, there were available plots which had been inoculated at 
seeding time, other plots which had been both inoculated and Hmed, 
and untreated check plots. Accordingly, three typical square yards 
in each of these three kinds of material were harvested, weighed, and 
analyzed. The number of plants and the total yield of tops and 
roots (both green weight and dry matter) for each of these nine 
square-yard tracts and the average total yield of dry matter per 
square yard for each of the three treatments are shown in Table 3. 

Table 3. — Effect of inoculation on growth of tops and roots of szveet clover 
and of alfalfa, as shown by the green and dry weights per square yard and the 
dry matter per plant {all weights in grams). 



SWEET CLOVER. 



Treatment. 


13: 


No. of 


Green weight. 


Total dry 
matter. 


Dry matter per plant. 


No. 


plants. 


Tops. 


Roots. 


Tops. 


Roots. 


Tops. 


Roots. 


Total. 


Average 
for plot. 


Inoculated and limed. 


I 


70 


2,062.5 


336.4 


534.5 


88.0 


7.636 


1.257 


8.893 








2 


82 


1,922.5 


291.5 


456.5 


70.3 


5-567 


0.857 


6.424 




6.796 




3 


108 


2,003.5 


271.0 


484-5 


63-2 


4.486 


0.585 


5-071 






Inoculated only 


4 


70 


1,216.7 


228.1 


301.7 


60.1 


4.310 


0.859 


5.169 








5 


57 


1,204.0 


194-5 


273.0 


45-1 


4.790 


0.791 


5-581 




■5-825 




6 


72 


1,828.0 


267.5 


416.0 


68.1 


5-778 


0.946 


6.724 








7 


80 


585-0 


134-7 


148.0 


36.0 


1,850 


0.450 


2.300 








8 


80 


406.3 


103.8 


97-3 


26.4 


1. 216 


0.330 


1.546 




1.799 




9 


79 


417-5 


95-0 


100.5 


22.0 


1.272 


0.279 


I-551 







ALFALFA. 





4 


139 


1,384 


363 


347-4 


107.0 


2.500 


0.769 


3-269 








5 


288 


1,789 


667 


469.4 


193-7 


1.629 


0.673 


2.302 




^ 2.789 




6 


169 


1,435 


461 


348.1 


124. 1 


2.060 


0.735 


2.795 






Not inoculated 


I 


157 


317 


188 


86.0 


60.0 


0.548 


0.382 


0.930 








2 


203 


213 


124 


57-8 


40.1 


0.285 


0.197 


0.482 




^ 0.786 




3 


140 


281 


147 


80.3 


52.1 


0-574 


0.372 


0.946 







These data show the same general effects from the inoculation as 
were found in the first year. The yields of dry matter per plant and 
per plot were greater on the inoculated and limed plot than on the 
plot which was inoculated but not limed, and on each of these the 
yields were more than three times as great as on the uninoculated 

» Loc. cit., p. 180-182. 



ARNV AND TirATC lIl'.R : INOCULATION OF ALFALFA. I33 

plot. This shows a very remarkable increase in growth in the first 
crop (second season's f^rowth) as a resnlt of the treatment at seeding 
time. 

Table 3 also gives data concerning the yields from the three square 
yards each of inoculated and uninoculated alfalfa. Here again the 
effect of the inoculation at seeding time upon the amount of crop 
produced the following season is very marked, the average yield of 
dry matter from the three inoculated square yards being almost four 
times as great as that from the uninoculated plots. 

As would be expected, the yield of dry matter per plant on the 
different square-yard plots varies inversely with the number of 
plants on each particular plot. There seems to be no exact correla- 
tion between the yield of dry matter per plant or per plot and the 
number of plants on each particular plot. In general, it appears 
that on the inoculated plots the thicker stand (i. e., larger number 
of plants per square yard) gave heavier yields of total dry matter, 
while on the uninoculated plots the thinner stands usually produced 
the greater growths of dry matter. There are several exceptions to 
this general rule, however, in the results of both seasons. 

The same effect of inoculation in increasing the proportion of tops 
to roots in both sweet clover and alfalfa that was noticed in the 
former work appeared in the data shown in Table 3. The ratios of 
dry matter in the tops to that in the roots is presented in Table 4. 



Table 4. — Effect of inoculation on ratio of tops to roots in sweet clover and 

' alfalfa. 



Crop. 


Treatment. 


Ratio of dry matter in tops to roots. 


.Square yard No. 


Average. 


I 


2 


3 


Sweet clover. . . . 
Alfalfa 


Inoculated and limed 

Inoculated only 

None 

Not inoculated 


6.07:1 
5.02:1 
4.11 :i 
3.19:1 
1.43:1 


6.49:1 
6.05:1 
3.68:1 
2.37:1 
1.44:1 


7.67:1 

6.II :i 
4-57:1 
2.80:1 
1.54:1 


6.74:1 

5-73:1 
4.09:1 
2.79:1 
1.48:1 



The data in Table 4 confirm those of the preceding year. In every 
individual case, of the total of 27 square yards which were investi- 
gated in the two seasons, the ratio of tops to roots was significantly 
greater in the plants from the inoculated plots than in those from 
plots which had not been inoculated. The same general effect is to 
be observed in the data presented by Fred and Graul in their recent 
paper^^ on the effect of soluble nitrogenous salts on nodule forma- 

10 Fred, E. B., and Graul, E. J. The effect of soluble nitrogenous salts on 
nodule formation. Jour. Amer. Soc. Agron., 8: 316-328. 1916. 



I 34 JOURNAL OF THE AMERICAN SOCIETY AGRONOMY. 

tion. In nearly every case reported by them the ratio of dry matter 
in tops to that in roots is greater in the pot where inoctdation was 
successful (as indicated by the formation of numerous nodules) than 
in the uninoculated pot which received otherwise the same treatment, 
this being true of both the alfalfa and the vetch crops. Their pot 
tests show occasional exceptions to the general rule, whereas our 
square-yard field tests exhibit no exceptions. The ratios vary with 
different square yards, as might be expected, but in no single case is 
the ratio of tops to roots as high in the uninoculated material as in 
that which received inoculation. 

The obvious conclusion from these results is that the effect of 
inoculation in increasing the yield of harvestable forage is due not 
only to an increase in the total dry matter per plant (as shown in 
Tables 4 and 5), but also to the deposition of a larger proportion of 
the material elaborated by each plant in the above-ground portion 
of it. 

The carefully prepared samples from these square-yard plots 
afforded excellent material for a continuation of the study of the effect 
of inoculation upon the quantity of plant- food constituents removed 
from the soil by these crops, which was begun last year. The samples 
were prepared and analyzed in the same way as described in the 
former paper, with the single exception that the percentage of calcium 
in the composite sample from each plot was determined, in addition 
to those previously reported. The results of the analyses are showti 
in Table 5. 

These data clearly show the following effects of the inoculation 
upon the percentages of plant food constituents in the crop : 

1. The percentage of ash in the dry matter is decreased in the 
tops and not affected or very slightly increased in the roots. 

2. The percentage of nitrogen is increased in both the tops ami the 
roots. 

3. The percentage of phosphorus is lower in both the tops and tfie 
roots of the inoculated sweet clover, but in the alfalfa there is na 
significant effect of the inoculation, the percentage of P2O5 being 
slightly lower in the tops and slightly higher in the roots of the 
inoculated plants. 

4. The percentage of potassium is higher in both the tops and the 
roots of the inoculated sweet clover, and lower in both the tops and 
the roots of the inoculated alfalfa. 

5. The percentage of lime in the dry matter does not appear to be 
significantly affected by inoculation. In the sweet clover samples, the 



ARNY AND TIIATCIIEU: INOCULATION OK ALFALFA. 



lime content of the incKulatccl and iininoculatcd plants is ]>ractically 
identical, while the plants from the plots which were limed at seed- 
ing time contain a lower percentage of calcium oxide than those from 
the unlimed plots. In the alfalfa samples, the percentage of lime is 
lower in the tops and higher in the roots of the uninoculated plants. 

Table 5. — Effect of inoculation on plant-food constituents in tops and roots of 
sweet clover and alfalfa. 



SWEET CLOVER. 



Treatment. 


Sq. 
Yd. 
No. 


Plant-food constituents (calculated as percentages of the 
dry matter.) 


Tops. 


Roots. 


Ash. 


Nitro 
gen. 




P206. 


K2O. 


CaO. 


Ash. 


Nitro- 
gen. 




P2O5. 


K2O. 


CaO. 


Inoculated and limed 


I 


6.81 


2.48 " 










4-54 


2.03 












2 


6.82 


2.54 




0.82 


2.49 


0.45 


5-39 


2.10 




0.93 


1.89 


0.07 




3 


7-15 


2.47 , 










5-15 


1.98 ^ 










Inoculated only 


A 


6.21 


2.30 










5.58 


1.96 












5 


6.97 


2.33 




0.69 


2.53 


0.58 


5-43 


2.18 




0.91 


1.80 


O.IO 




6 


6.41 


2.24 










4-95 


1.89, 












7 


7-35 


I-3I ' 










4.62 


0.89 












8 


7-77 


1.44 




1. 12 


1.78 


0.60 


4-93 


0.95 




1.02 


1.68 


O.IO 




9 


7-49 


1-35 ; 










4-83 


0.82 











ALFALFA. 





4 


8.33 


2.59 










3-98 


1.94 ^ 












5 


7.74 


2.36 




• 77 


1. 12 


1.69 


4-58 


2.20 




.82 


.65 


.19 




6 


8.32 


2.74 , 










4-45 


2.29 










Not inoculated. . 


I 


9-65 


1.56 










4.09 


• 73 ' 












2 


10.52 


1-59 




.89 


1.74 


2.17 


4-74 


• 75 




•70 


.88 


•13 




3 


10.18 


1-37 . 










3-50 


•65 J 











The percentages of these plant-food constituents in the dry matter 
are, of course, more or less dependent upon the stimulating effect of 
the inoculation upon the growth of the plants. For example, an 
increased production of protein in the plant results in a diminished 
percentage of all these plant- food constituents except nitrogen, while 
a stimulation of carbohydrate production would decrease the per- 
centage of all plant-food elements in the dry matter. For this reason, 
a better basis of study of the resultant effect of inoculation upon the 
ability of the plant to draw upon these plant food elements in the soil 
is the total amount of these elements found in the dry matter from 
each identical area. These computations have therefore been made, 
and the resulting data are presented in Table 6. 

These data clearly show that inoculation, resulting in increased 
growth, gives to the crop a very largely increased power to draw upon 
the plant- food constituents of the soil. The effect does not seem to 



136 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

be simply the result of increased growth processes, since it is not 
uniform for the several mineral elements, but is generally greater 
in the case of potassium than of the other elements which were de- 
termined in these analyses. The figures for total ash in the dry 
matter in both seasons indicate that the inoculated plants are able to 
elaborate a greater amount of organic material from a given amount 
of mineral constituents than are the uninoculated ones. There seems, 
therefore, to be some physiological effect upon the growth of the 
plants, due to the presence of nodule-forming bacteria on their roots, 
other than a simple increase in nitrogen supply to the growing plants. 

Table 6. — Effect of inoculation upon quantity of plant food constituents in 
crops of sweet clover and alfalfa. 

(Average of three square yard plots in each case.) 



Plant-food constituents present in crop. 



Crop and treatment. 


Part of 


plants. 


Grams per square yard. 


P 


ounds per acre 






Nitro- 
gen. 


P2O5. 


K2O. 


CaO. 


Nitro- 
gen. 


P2O5. 


K2O. 


CaO. 


Sweet clover: 






















Inoculated and limed. . 


Tops 




12.21 


4-03 


12.25 


2.21 












Roots 




1.50 


0.68 


1.39 


0.05 












Whole 


plant 


13-71 


4.71 


13.64 


2.26 


153 


52 


152 


25 




Tops 




7-54 


2.28 


8.35 


I.9I 












Roots 




1. 12 


0.53 


1.04 


0.06 












Whole 


plant 


8.66 


2.81 


9-39 


1.97 


96 


31 


105 


22 




Tops 




1.56 


1.29 


2.05 


0.69 












Roots 




0.25 


0.29 


0.47 


0.03 












Whole 


plant 


1. 81 


1.58 


2.52 


0.72 


20 


18 


29 


8 


Alfalfa: 
























Tops 




9.87 


2.99 


4-35 


7.16 












Roots 




3-o6 


1. 16 


0.92 


0.27 












Whole 


plant 


12.93 


4-15 


5.27 


7.43 


134 


46 


59 


83 


Not inoculated 


Tops 




1. 12 


0.66 


1.30 


1.62 












Roots 




0.35 


0.36 


0.45 


0.07 












Whole 


plant 


1.47 


1.02 


1-75 


1.69 


16 


II 


19 


19 



Summary and Conclusions. 

Studies have been carried on in two successive seasons (1914 and 
191 5) of the effect of inoculation by various methods at seeding 
time upon the yield and composition of the crops grown in the 
second, third, and fourth seasons thereafter, on three different fields 
of the University Farm, St. Paul, Minn. 

In this and in the preceding paper of the same title, data are pre- 
sented which show (a) the yield and protein content of the three 
successive cuttings in each of the two years from fourteen inoculated 
plots and nine adjacent uninoculated check plots of alfalfa; {h) the 



ARNY AND THATC IIKU: INOCUf.ATION OF ALFALFA. I 

yield and protein content of one year's crop from two inoculated and 
three uninoculated sweet clover plots; and (c) two successive seasons' 
analyses of the tops and roots from 3 square yards each of inoculated 
and uninoculated alfalfa and sweet clover. 

These data point clearly to the following conclusions with refer- 
ence to the effect of inoculation upon the yield and composition of 
these leguminous crops when grown on soils of the type represented 
by those on University Farm. 

1. Inoculation at seeding time produces a very large increase in 
yield of dry matter per acre and in the percentage of protein in the 
dry matter in the second season thereafter (first harvestable crop), 
as compared with the yield and composition of the crop from ad- 
jacent uninoculated plots. In the next season's growth (second 
harvestable crop), the differences are much less noticeable and they 
practically disappear in the following year, by reason of the rapid 
spread of the inoculating bacteria to the uninoculated plots. 

2. Inoculation of either alfalfa or sweet clover with soil from 
fields on which either alfalfa or sweet clover has been growing suc- 
cessfully is equally efffcient in producing these effects. 

3. Inoculation with soil is generally more efficient in these respects 
than inoculation with the commercial cultures which were used in our 
experiments. 

4. Liming the soil at seeding time (2 tons ground limestone per 
acre) slightly intensifies the above-mentioned effects of inoculation. 

5. One effect of inoculation is to give to the inoculated plants an 
increased capacity to utilize mineral soil nutrients, the increased 
growth resulting in the removal from the soil of very much larger 
amounts of potassium, phosphorus, and calcium. 

6. A second effect of inoculation is to make it possible for the 
inoculated plants to elaborate a somewhat larger amount of dry 
matter from a given amount of mineral plant food elements. 

Minnesota Agricultural Experiment Station, 
University Farm, St. Paul, Minn. 



138 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



A NEW METHOD FOR HARVESTING SMALL GRAIN AND 

GRASS PLOTS. 

A. G. McCall. 

In varietal testing and in soil fertility plot work one of the most 
serious difficulties encountered is that of obtaining accurate yields of 
the several plots included in the test. While the harvesting and weigh- 
ing of the entire product of the plot should, theoretically, be most 
satisfactory, this method is attended by certain difficulties that make 
it practically impossible to get accurate results. The lodging of the 
grain, the depredations of birds and mice, and losses incident to weath- 
ering are some of the more serious difficulties met when the entire plot 
is harvested. When plot work is being conducted in several sections 
of the State another difficulty is encountered, namely, the supplying 
of suitable machinery to thrash the grain from the small plots sepa- 
rately. 

These difficulties have been overcome to a certain extent by har- 
vesting a number of small areas from each plot and calculating from 
these the yield of the entire plot. The chief objections to this method 
are (i) the inconvenience encountered in laying off the small areas, 
and (2) the difficulty of obtaining representative areas. 

To overcome the first objection the writer has recently constructed, 
for the work at the Maryland Agricultural Experiment Station, a 
small apparatus to be used in harvesting accurate areas of grass, 
wheat, or other small grains. The essential details of this harvester 
are shown in figure 7. For the harvesting of grain that has been 
seeded in drills only the grain board (A) and the two spears (B, B') 
are used, the method of procedure being as follows : With the spears 
withdrawn the grain board is placed in position parallel and close to 
the outside drill row of the plot and fastened in this position by 
means of two short spears (not shown in the drawing) which project 
into the soil from the lower edge of the board. The long spears, 
B, B\ are now thrust through the metal sleeves at C, C and into the 
plot, thus marking off two parallel lines at right angles to the drill 
rows. The distance between these spears should be such that 4 or 5 
drill rows thus marked off will give an even fraction of an acre. 

For grass and for grains that are seeded broadcast the manipula- 
tion is essentially the same as for the drilled grain except that, since 



MCCALL: TIARVESTFNtl GRASS AND r.KAIN PLOTS. 



there is no drill row, the fourth side of the area must he defined by the 
use of a third spear, which nuist be parallel to the <^rain board. To 




Fig. 7. Frame for use in harvesting small blocks from grain and grass plots. 



accomplish this the two spears, B, B' , are securely clamped in position 
by means of the bar D, after which the sleeve F is clamped to the 
spear B' near the tip and at a definite distance from the board A. 
The third spear, E, is now thrust through the sleeve F until it engages 
the sleeve G which is clamped to spear B directly opposite the sleeve 



PLOT 



i a"| [a] : A i 

Fig. 8. Diagram showing location of blocks (A) in grain or grass plots in 
order to obtain representative samples. 

F, thus marking ofif the fourth side of the small area that is to be har- 
vested. Having defined the boundaries it only remains to cut the 
grain or the grass included within the spears. 

By the aid of this apparatus it is possible to measure with great 



140 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

accuracy an area of small dimensions even though the grain or grass 
is badly lodged and tangled. It is proposed to harvest 5 of these small 
areas of one five-thousandth of an acre each, or a total of one one- 
thousandth of an acre, from each large plot and to thrash the grain 
with a small thrashing outfit. The locations of these small areas are 
shown in figure 8. The small bundles of grain are taken to the barn 
the same day they are harvested, suspended on drying racks until 
thrashed, and weighed when properly cured. Plots in remote parts 
of the State will be harvested in a similar manner and the bundles 
of grain shipped to the central station in canvas bags. 

A preliminary trial of the apparatus on wheat and on timothy plots 
gave results that check up quite satisfactorily with the records ob- 
tained by harvesting and thrashing the entire plots. 

This brief account is presented with the hope that some members of 
the American Society of Agronomy will be sufficiently interested to 
construct a similar apparatus, use it during the next harvest period, 
and report their experience at the next meeting of the Society. 

Maryland Agricultural Experiment Station, 
College Park, Maryland. 



AGRONOMIC Al' FAIRS. 



141 



AGRONOMIC AFFAIRS. 

MEMBERSHIP DUES. 

All those whose membership dues for 191 6 in the American Society 
of Agronomy are not paid by March 31 will automatically lapse. 
The names of lapsed members will be printed in the May number of 
the Journal. The few whose dues are still unpaid are urged to 
remit promptly in order that they may be reinstated before that time. 

By reason of the change in the by-laws of the Society effected at 
the last annual meeting, those whose dues for 191 7 are not paid by 
April I will receive no more numbers of the Journal until they remit. 
This provision was made in order to comply with the postal laws and 
also to safeguard those members whose dues are paid promptly. One 
who delays the payment of annual dues until several notices have 
been sent him entails unnecessary expense on the Society and causes the 
Treasurer needless work. One who does not pay at all should not, as 
has been the case in the past, receive the Journal for more than a 
year without cost. All members are urged to remit their dues for 
191 7 to the Treasurer so that there will be no delay in the delivery 
of the Journal to them. 

MEMBERSHIP CHANGES. 

The membership of the Society as reported in the January Journal 
was 618. Since that time 31 new members have been added and 7 
members have resigned, a net gain of 24 and a present membership 
of 642. The names and addresses of the new members, the names 
of those who have resigned, and such changes of address as have 
come to the notice of the Secretary follow. 

New Members. 

Agee, John H., Bur. Soils, U. S. Dept. Agr., Washington, D. C. 
Bell, N. Eric, Agriculture & Industries Dept., Montgomery, Ala. 
BiNFORD, E. E., Substation No. i, Beeville, Tex. 

Brandon, Joseph F., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Bryant, Roy, hi West Street, Stillwater, Okla. 

Clemmer, H. J., Woodward Field Station, Woodward, Okla. 

CoE, H. S., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C 

Conrey, G. W., Soils Building, University of Wisconsin, Madison, Wis. 



142 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Davisson, Bert S., Ohio Agr. Expt. Sta., Wooster, Ohio. 

Deeter, E. B., Bur. Soils, U. S. Dept. Agr., Washington, D. C. 

Downs, E. E., Michigan Agr. College, East Lansing, Mich. 

Fletcher, C. C, Bur. Soils, U. S. Dept. Agr., Washington, D. C. 

Fletcher, O. S., Ellensburg High School, Ellensburg, Wash. 

Furry, R. L., Ferguson Seed Co., Sherman, Tex, 

HoTCHKiss, W. S., Substation No. 2, Troup, Tex. 

Huelskemper, Edward H., 411 Terry Street, Longmont, Colo. 

Hurst, J. B., in Knoblock Street, Stillwater, Okla. 

Jarrell, J. F., Expt. Dept., Great Western Sugar Co., Longmont, Colo. 

KusKA, J. B., Colby Substation, Colby, Kans. 

McIlvaine, T. C, W. Va. Agr. Expt. Sta., Morgantown, W. Va. 

McNess, Geo. T., Nacogdoches Substation, Nacogdoches, Tex. 

MuNDELL, J. E., Big Springs Expt. Farm, Big Springs, Tex. 

Pate, W. F., N. C. Agr. Expt. Sta., West Raleigh, N. C. 

Reed, Everett P., N. Y. Agr. Expt. Sta., Geneva, N. Y. 

ScHUER, Henry W., 296 Chestnut St., Chillicothe, Ohio. 

Shinn, E. H., 238 Main Street, Stillwater, Okla. 

Smith, J. B., Dept. of Farm Crops, College of Agr., Columbia, Mo. 

Tillman, B. W., Bur. Soils, U. S. Dept. Agr., Washington, D. C. 

True, Rodney H., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Ward, Wylie R., University Farm, Lincoln, Nebr. 

Woo, Moi Lee, Spreckels, Cal. 



Members Resigned. 

Buell, T. W. Holtz, Henry F, Nuckols, S. B. 

Cory, Victor L. LeClair, C. A. Orton, W. A. 

McMuRDo, Geo. A. 



Addresses Changed, 

DoRSEY, Henry, 306 Stewart Ave., Ithaca, N. Y. 

Garland, John J., Holmes-Letterman Seed Co., Canton, Ohio. 

Hendry, Geo. W., University Farm, Davis, Cal. 

Hershberger, Jos. P., jr., 1835 Indianola St., Columbus, Ohio. 

Hill, W. H., Laboratory Inland Rev. Dept., 317 Queen St., Ottawa, Ont. 

Martin, Thos. L., Castle Dale, Utah. 

Phillips, Thos. G, College of Agriculture, Columbus, Ohio. 

TuTTLE, H. Foley, Dept. Soil Physics, University of Illinois, Urbana, III 

Welch, J. S., Paradise, Utah. 



AGRO N O M I C A I'- !• A I H S . 



NOTES AND NEWS. 

A. C. Hartenbower has resigned as agronomist in charge of the 
Guam station and has returned to the continental United States. He 
has been succeeded in Guam by C. W. Edwards, formerly engaged 
in agricultural work in the Philippines. 

R. B. Lowry succeeded J. C. Pridmore as associate professor of 
agronomy in the University of Tennessee on February i. Mr. Prid- 
more's resignation to enter commercial work was previously noted. 

George E. Vincent, for the past several years president of the Uni- 
versity of Minnesota, has resigned to become president of the Rocke- 
feller Foundation and has entered on his duties in the latter position. 
Marion L. Burton, president of Smith College, has been elected to 
succeed him in Minnesota and has accepted the position. 

J. S. Welch, superintendent of the Gooding (Idaho) substation, 
has resigned and has sailed for New Zealand, where he will teach 
agriculture in the Maori agricultural college. 

Albert F. Woods, dean of the Minnesota college of agriculture and 
director of the experiment station since 1910, has accepted the presi- 
dency of the Maryland State College and will begin his new duties 
on July I. His new position entails the supervision of all the agricul- 
tural activities of the State. 

Several administrative changes have been effected in the Federal 
Ofhce of Dry-Land Agriculture. John S. Cole will have general 
charge of field stations, assisted by John M. Stephens in the northern, 
O. J. Grace in the central, and E. F. Chilcott in the southern Great 
Plains. Messrs. Stephens, Grace, and Chilcott will retain the super- 
intendencies of the Judith Basin (Mont.), Akron (Colo.), and Wood- 
ward (Okla.) field stations, respectively. W. E. Lyness has been 
transferred from the Akron to the Archer, Wyo., station, and has 
been succeeded at Akron by Jos. F. Brandon of the University of 
Illinois. Albert Osenbrug has been assigned to the Scottsbluff, 
Nebr., station, and H. J. Clemmer to the station at Woodward, Okla. 

J. A. Holden has succeeded Fritz Knorr as superintendent of the 
Scottsbluff, Nebr., experiment farm of the Office of Western Irriga- 
tion Agriculture. Mr. Knorr is now engaged in farming in northern 
Alabama. 



144 journal of the american society of agronomy. 

Local Sections. 

The fifeenth regular meeting of the Washington (D. C.) section 
was held at the Cosmos Club, February 9, 191 7. The program con- 
sisted of a paper entitled " Some Features of Soil Classification," by 
Dr. C. F. Marbut, in charge of soil survey in the U. S. Department 
of Agriculture. Dr. Marbut defined soil from the standpoints of the 
geologist, pedologist, and agronomist, stating that to the latter a soil 
is anything in which plants grow. He showed that soils derived from 
the same or similar rocks may vary widely, while similar soils may 
be derived from very different rocks, illustrating the futility of a 
geologic classification of soils. He then discussed at length the sev- 
eral soil belts which extend around the world, as pointed out in a 
Russian work on soils published in 191 4 and which is still extremely 
rare in this country. According to the Russian scientists, there is a 
great belt of tundra in the north, the progression from north to south 
being through belts of gray forest and black prairie soils into the 
yellow and red soils of the semitropical and tropical countries. 

The sixteenth regular meeting of the Washington section was held 
on March i . The program included papers on " Experiments on the 
Effect of Fall Irrigation on Crop Yields," by F. D. Farrell; "Pro- 
gression of the Wheat Harvest from South to North in the Great 
Plains," by Joseph F. Brandon ; and a report of recent studies in 
evaporation and transpiration, by L. J. Briggs and H. L. Shantz. 

The first regular meeting of the Columbus (Ohio) section during 
1917 was held February i, during Farmers' Week. Dr. E. R. Allen 
of the Ohio station presented an interesting paper on " The Attack on 
the Problem of Soil Biology." The meeting was so enjoyable and 
profitable that it was decided to continue the practice of holding a 
meeting of the section during Farmers' Week, making it the one big 
affair of the year. 

The following papers have been presented before the South Dakota 
section at Brookings since January i : 

On January 12, "Wheat Rust," by A. N. Hume; on February 2, 
" Transpiration of Sweet Clover in Different Soil Types," by Howard 
Loomis; and on March i, "Cooperative Cereal Experiments at High- 
more " and " Sorghum Culture in South Dakota," by J. D. Morrison 
and George Winwright. 



JOURNAL 

OF THE 

American Society of Agronomy 



Vol. 9. April, 191 7. No. 4. 



THE EFFECT OF SODIUM NITRATE APPLIED AT DIFFERENT 
STAGES OF GROWTH ON THE YIELD, COMPOSITION, 
AND QUALITY OF WHEAT.^ 

J. Davidson and J. A. Le Clerc. 
Introduction. 

The work of Le Clerc and Leavitt^ showed that the variation in 
nitrogen content of wheat is independent of the original nitrogen 
content of the seed used. This conclusion has since been corroborated 
by data which have not yet been published with reference to wheat 
and other cereals. The work of Shaw and Walters^ and also that of 
Le Clerc and Yoder^ further showed that the soil is a minor factor 
in accounting for the variation in the nitrogen content of wheat. 
The logical conclusion of these experiments is that the principal 
factor producing variations in the nitrogen content of wheat within 
the limits of these experiments is climate. Climate, however, is a 
complex of a number of factors, as rainfall, sunshine, elevation, 
minimum and maximum temperature, etc. It remains to be found 

1 Contribution from the Plant Chemical Laboratory of the Bureau of Chem- 
istry, U. S. Department of Agriculture. Presented at the ninth annual meeting 
of the American Society of Agronomy, Washington, D. C, November 14, 1916. 

2 Le Clerc, J. A., and Leavitt, Sherman. Trilocal experiments on the influ- 
ence of environment on the composition of wheat. U. S. Dept. Agr., Bur. 
Chem. Bui. 128, 18 p. 1910. 

3 Shaw, Geo. W., and Walters, E. C. A progress report upon soil and cli- 
matic factors influencing the composition of wheat. Cal. Agr. Expt. Sta. Bui. 
216, p. 54^574- 1911- 

4 Le Clerc, J. A., and Yoder, P. A. Environmental influences on the physical 
and chemical characteristics of wheat. U. S. Dept. Agr., Jour. Agr. Research, 
v. I, no. 4, p. 275-291. 1914. 

145 



146 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

out which factor or which combination of factors composing climate 
is responsible for these variations. It remains further to be de- 
termined whether the action of climate is direct, affecting the meta- 
bolism of the plant, or if it is indirect, affecting the amount of avail- 
able plant food in the soil. The work reported here was undertaken 
as the first step in answering these questions. It was thought that 
perhaps climate was responsible for the variation in the available 
nitrates at different stages of growth, and it was therefore deemed 
advisable to see what would be the effect of applying sodium nitrate 
at different stages of growth. 

Plan of Experiments. 

Size of Plot. — There has been much discussion in agronomic litera- 
ture with reference to size of plot for field experiments. The tenth- 
acre plot is commonly accepted. The reason for using such large 
plots is to overcome the lack of uniformity in the soil and to allow 
for variation in the vitality of the seed used. In the present experi- 
ments there has been a conspicuous deviation in the method of laying 
out and in the size of the plots from the general practice. The plots 
were laid out after the crop was up. Areas showing uniform growth 
were selected and therefore it was possible to limit the plots to i 
square rod each. The results, as will be shown further, have proved 
quite satisfactory, and would no doubt have been much more satis- 
factory if the land had been more carefully prepared and drilled. 

Method of Applying Fertilizer. — The sodium nitrate was used at 
the rate of 2 pounds per plot, equivalent to 320 pounds per acre. It 
was applied in one period, two periods, and three periods. The 
periods chosen were : 

(1) When the crop was about 2 inches high. 

(2) Time of heading. 

(3) Milk stage. 

In order to assure the availability of the fertilizer at the particular 
stages of growth, the nitrate was applied in solution. The concentra- 
tion of the solution was i to 100 in all cases. 

To check the advisability of applying the fertilizer in solution in 
the future, parallel plots were prepared to which the solid fertilizer 
was applied. These plots received the same amount of water as the 
plots to which the fertilizer was applied in solution, except that the 
water was added the day before the application of the nitrate. The 
plots which received the fertilizer in two or three periods also re- 
ceived the same amount of water as those plots which received the 



DAVIDSON \- LK C I.I-.KC : l-.KKKCT Ol*' NITRATl-: ON WIIi:\T. I47 



fertilizer in one period. The additional water above the amount 
ro(|uired to make up the proper concentration was added the previous 
(lay. Therefore, all these plots received the same amount of water 
at each of the three different stages of growth. 

To check the effect of water a series of plots was prepared to 
which the fertilizer was applied in solid state and to which no water 
was added. An additional scries of plots was prepared to which 
sodium nitrate and potassium chloride were applied, and also one to 
which potassium chloride alone was added. These series were planned 
in order to see whether the relation between nitrates and potash 
on the one hand and the quality and composition of wheat on the 
other found by Headden^ would hold good under these conditions. 
The experiment was carried on in duplicate series. The details of the 
treatment which was given the individual plots are presented in the 
following pages. The experiment was carried out in 191 6 on the 
Kentucky Agricultural Experiment Station Farm at Lexington, Ky.^ 



Treatment of Individual Plots at Different Stages of Growth. 



Plot 
No. 



Time of heading. 

25 gallons of water. 



Do. 



Milk stage. 

25 gallons of water. 
Do. 



When 2 inches high. 

1. 2 pounds of NaNOs 

dissolved in 25 gal- 
lons of water. 

2. 2 pounds of NaNOs 

and 25 gallons of 
water the day pre- 
vious to the appli- 
cation of fertilizer. 

3. % pound of NaNOs 

dissolved in SYo gal- 
lons of water and 
iGYo gallons of wa- 
ter the day previous 
to the application of 
fertilizer. 

4. pound of NaNOs 
and 25 gallons of 
water the day pre- 
vious to application 
of fertilizer, 

5. 25 gallons of water. 

5 Headden, W. P. Colo. Agr. Expt. Sta. Bui. 205. 

6 Occasion is hereby taken to express our gratitude to the late Dr. J. H. 
Kastle, director of the Kentucky Agricultural Experiment Station, for his cour- 
tesy in offering us the facilities of the station, and to Prof. E. J. Kinney, of 
the same station, for his assistance in carrying out the experiment. 



% pound of NaNOs dis- 
solved in 8V2 gallons 
of water, and 16^/2 
gallons of water the 
day previous to the 
application of the fer- 
tilizer. 

% pound of NaNOs and % pound of NaNOs and 
25 gallons of water 25 gallons of water 



% pound of NaNOs dis- 
solved in 8^/^ gallons 
of water, and 16"!,^ 
gallons of water the 
day previous to the 
application of the fer- 
tilizer. 



the day previous to 
the application of the 
fertilizer. 
25 gallons of water. 



the day previous to 
the application of the 
fertilizer. 
25 gallons of water. 



14^ JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



6. I pound of NaNOs 

dissolved in 12^^ 
gallons of water 
and I2y2 gallons of 
water previous to 
the application of 
fertilizer. 

7. I pound of NaNOs and 

25 gallons of water 
the day previous to 
the application of 
fertilizer. 

8. 25 gallons of water. 



9. Do. 



10. I pound of NaNOs 

dissolved in 12% 
gallons of water 
and 121/^ gallons of 
water the day pre- 
vious. 

11. I pound of NaNOs 

and 25 gallons of 
water the day pre- 
vious to the appli- 
cation of fertilizer. 

12. 25 gallons of water. 

13. Do. 



14. Do. 



15. Do. 



I pound of NaNOs dis- 
solved in I2i/^ gallons 
of water and I2y2 
gallons of water the 
day previous to' the 
application of the fer- 
tilizer. 

I pound of NaNOs and 
25 gallons of water the 
day previous to the 
application of fertil- 
izer. 

I pound of NaNOs dis- 
solved in 12^/^ gallons 
of water and 12^,^ 
gallons of water the 
day previous to the 
application of fertil- 
izer. 

I pound of NaNOs and 
25 gallons of water 
the day previous to 
the application of fer- 
tilizer. 

25 gallons of water. 



Do. 



2 pounds of NaNOs dis- 
solved in 25 gallons of 
water. 

2 pounds of NaNOs and 
25 gallons of water 
the day previous to 
the application of fer- 
tilizer. 

25 gallons of water. 



Do. 



25 gallons of water. 



Do. 



I pound of NaNOs dis- 
solved in I2i/^ gallons 
of water and 12^/^ 
gallons of water the 
day previous to the 
application of fertil- 
izer. 

I pound of NaNOs and 
25 gallons of water 
the day previous to 
the application of fer- 
tilizer. 

I pound of NaNOs dis- 
solved in 12^/^ gallons 
of water and i2-^ 
gallons of water the 
day previous. 

I pound of NaNOs and 
25 gallons of water 
the day previous to 
the application of fer- 
tilizer. 

25 gallons of water. 

Do. 



2 pounds of NaNOs dis- 
solved in 25 gallons of 
water. 

2 pounds of NaNOs and 
25 gallons of water 
the da}" previous to 



DAVIDSON \ LK CLKRC I KI FKCT OK NITRATE ON WHEAT. 1 49 



16. 2 pounds of NaNO,. 
17- % pound of NaNO;,. 
18. I pound of NaNOs. 
19: No treatment. 

20. I pound of NaNOs 

21. No treatment. 

22. Do. 

23. Do. 

24. 2 pounds of NaNOs 

and 2 pounds of 
KCl dissolved in 25 
gallons of water. 

25. 25 gallons of water. 



26. Do. 



27. 2 pounds of KCl dis- 

.<:olved in 25 gallons 
of water. 

28. 25 gallons of water. 



29. Do. 

30. 2 pounds of NaNOs 

and 2 pounds of 
KCl. 

31. No treatment. 

32. Do. 

33. 2 pounds of KCl. 

34. No treatment. 
35- Do. 



No treatment. 

% pound of NaNOa. 

1 pound of NaNOa. 
Do. 

No treatment. 

2 pounds of NaNOs. 
No treatment. 

Do. 

25 gallojis of water. 



Do. 



2 pounds of KCl dis- 
solved in 25 gallons of 
water. 

25 gaflons of water. 



No treatment. 



Do. 

2 pounds of KCl. 
No treatment. 



the application of fcr 

tilizcr. 
No treatment. 
% pound of NaNOs. 
No treatment. 

1 pound of NaNOs. 
Do. 

No treatment. 

2 pounds of NaNOs. 
No treatment. 

25 gallons of water. 



2 pounds of NaNOs and Do 
2 pounds of KCl dis- 
solved in 25 gallons of 
water. 

25 gallons of water. 



2 pounds of NaNOs and 
2 pounds of KCl dis- 
solved in 25 gallons of 
water. 
25 gallons of water. 



Do. 



2 pounds of KCl dis- 
solved in 25 gallons of 
water. 

No treatment. 



2 pounds of NaNOs and Do. 

2 pounds of KCl. 
No treatment. 



2 pounds of NaNOs and 

2 pounds of KCl. 
No treatment. 
Do. 

2 pounds of KCl. 



Results. 

Yield and Percentage of Grain in Crop. — As seen from Table i, 
only those plots which received the nitrate at the first stage of growth 
responded in yield to the application of this fertilizer. Of the 70 
plots of the experiment 28 received varying amounts of sodium nitrate 
at the first stage, of which 10 received the nitrate at the rate of 320 



150 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



pounds per acre, 12 at the rate of 160 pounds per acre, and 6 at the 
rate of 106^ pounds per acre. All these plots showed a decided 
response in yield, the response being distinctly proportional to the 
amount of nitrate applied at this stage of growth. The nitrate ap- 
plied at the second and likewise at the third stage of growth did not 
seem to affect the vegetative growth of the crop in the slightest 
degree. 



Table i. — Yield and percentage of grain from plots to which nitrate of soda 
was applied in various quantities at various stages of growth. 



Fertilizer added at each 
application. 



Stages of growth. 



Yield of grain, pounds. 



Water applied. 



Ferti- | Ferti- 
lizer in ! lizer in 
solu- solid 
tion. state. 



No 
water 
applied. 



Percentage of grain in 
crop. 



Water applied. 



Ferti- 
lizer in 
solu- 
tion. 



Ferti- 
lizer in 
solid 
state. 



2 lbs. NaNOs . . 

Do 

Do 

1 lb. NaNo3 . . . 

Do 

Do 

2/3 lb. NaNOs. 

2 lbs. NaNOs + 
2 lbs. KCl 

Do 

Do 

2 lbs. KCl 

Do 

Do 

Check 



First 
Second 

Third 

First and second 

Second and third 

First and third 

First, second 
and third 

First 

Second 

Third 

First 

Second 

Third 



34-0 
28.8 
14.7 
14.8 
16.3 
13.2 
25.6 
27.2 
16.8 
12.2 
25.6 
25-5 
20.6 
21.2 
32.2 
37.1 
14.4 
19.2 
14.8 
17.6 
14.4 
17-3 
14.6 
18.1 
14-5 
15.6 



34-2 
26.4 
17.0 
14.6 
14.8 
13-9 
24.8 
24.4 
14.6 
14.2 
23.8 
23.8 
20.0 
21.0 



29.6 
28.8 
15-8 
II. 8 
16.3 
14.0 
25.0 
26.4 
15.2 
12.6 
25.0 
24.4 
22.5 
25-3 
32.1 
34-2 
15.2 
17.1 
13.0 
15-4 
16.6 
15-0 
15-0 
15-4 
12.4 

15-9 



18.8 
14.4 



13.0 
16.3 



34- 9 
30.0 

35- 9 
37.1 
36.8 
36.9 
36.4 
36.0 
35.0 
39-5 
33-3 
37-0 
38.6 
37-1 
37.8 
33.7 
37.8 
37.8 
37.3 
36.5 
39-0 
38.0 
28.2 
36.8 
37-5 
37-1 



The plots to which nitrate was added in solid form showed the 
same tendency as those which received it in solution. It will be 
noticed, however, that of the plots which received their nitrate at the 
first stage, those which received it in solution showed a tendency to 
yield somewhat higher than those which received it in solid form. A 
possible explanation of this phenomenon lies in the fact that the 



DAVins(^N & r-K CLKRi:: kkkkct of nitkatk on wheat. 151 

fertilizer was less evenly distributed in the plots to which it was 
applied in solid form. Certain spots in these plots probably received 
the nitrate at a rate which lies beyond the maximum point of its 
efficiency. This would naturally be at the expense of other spots in 
the same plots, which would have received the nitrate at a rate less 
than the nominal rate of application at this particular stage. 

No appreciable differences have been noted with reference to the 
effect of the fertilizer on the ratio between grain and straw. 

Percentage of Yellowberry and the Protein Content. — The terms 
"yellowberry " and **flinty " used in this paper refer only to the outer 
aspect of the grain, as the wheat used was a soft winter wheat with 
a starchy appearance inside. In determining the percentage of yellow- 
berry in a sample any kernel which had the slightest trace of the 
characteristic yellow coloration was classed as yellowberry. This ac- 
counts for the degree of irregularity which is observed in the figures 
representing the percentage of yellowberry in a few of the samples. 
This irregularity, however, does not mask the general tendency, which 
is shown very distinctly. This tendency is still more accentuated 
when the grain of the individual samples is examined in mass. 

As seen from Table 2, the samples grown on the plots which 
received an application of sodium nitrate in the second stage gave by 
far the highest protein content and the lowest percentage of yellow- 
berry. This is clearly shown by every one of the 28 plots which 
received varying application's of the nitrate at the second stage. The 
protein content and the flintiness of the grain grown on those plots 
vary directly with the quantity of fertilizer received at this stage. It 
is therefore the presence of nitrate at the second stage which affected 
the coloration and the protein content of the grain. The grain grown 
on the plots to which the nitrate was applied at the first or third 
stage showed but a slight increase in protein. 

It will be noticed that of the plots which received their nitrate in 
a single application those which received it at the third stage showed 
a tendency to give a somewhat higher protein content than those which 
received it at the first stage. It will be further noticed that of those 
plots which received their fertilizer in two applications those which 
received it at the second and third stages gave a distinctly higher 
protein content than those which received it at the first and second 
stages. It is possible that the third stage is more conducive to the 
formation of a somewhat higher nitrogen content in the grain than 
the first stage. It is also possible that the difference is due to the fact 
that the increased growth caused by the first application used up 



152 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



more plant food and thus left less of the nitrate in the soil to affect 
the protein content of the grain. 



Table 2. — Percentage of yellowberry and protein content of wheat from plots 
to which nitrate of soda was applied in various quantities at three 
stages of growth. 



Fertilizer added at each 
application.. 



Stages of growth. 



Percentage of yellqwherry. 



Water applied. 



Ferti- 
lizer in 
solu- 
tion. 



Ferti- 
lizer in 
solid 
state. 



No 
water 
applied. 



Percentage of protein. 



Water applied. 



Ferti 
lizer i 
solu- 
tion. 



Ferti- 
lizer in 
solid 
state. 



2 lbs. NaNOs . . 

Do. 

Do 

1 lb. NaNOs . . . 

Do 

Do 

2/3 lb. NaNOs. 

2 lbs. NaNOs + 
2 lbs. KCl 

Do 

Do 

2 lbs. KCl 

Do 

Do 

Check 7 



First 

Second 

Third 

First and second 

Second and third 

First and third 

First, second 
and third 

First 

Second 

Third 

First 

Second 

Third 



35 
40 
o 
o 
41 
36 
10 
10 
2 
I 

34 
47 
II 
18 
34 
27 
I 

- I 
40 
42 
71 
65 
68 

55 
70 
46 



28 

35 
o 
I 
35 
43 
9 
6 
2 
4 
33 
40 
19 
30 



23 
17 
I 
o 
^o 
37 
2 
10 
I 
3 
37 
38 
10 
15 
33 
27 
I 
I 
40 
40 

65 
62 

71 
47 
72 
48 



10.85 
10.95 
13-63 
14.04 
10.92 
II. II 

11-95 
11.87 
12.95 
13.50 
10.65 
10.75 
11.44 
11.44 

10.73 
11,24 

13-99 
14.14 
11.07 
II. II 
10.03 
9.98 
9.82 

9.94 
10.00 
9.84 



10.80 
10.80 
13-18 
13-50 
10.68 
10.70 
12.15 
12.13 
13.02 
13.60 
10.48 
10.94 
11.87 
11.80 



64 
68 



49 
40 



With reference to the coloration of the grain, there is some slight 
indication that those plots which received their nitrate in the first 
stage only gave a smaller percentage of yellowberry than those which 
received it at the third stage only. It is possible that it was due to 
the residual sodium nitrate left over in the soil from the first stage 
to be used by the plant at the second stage of growth. 

When potassium chloride was used alone in any plot the percentage 
of yellowberry in the grain seemed to be increased. 

Weight per Bushel and Weight per 1,000 Kernels. Table 3 shows 
that there is no consistent variation in the weight per 1,000 kernels 
and in the weight per bushel. The variations are slight and do not 



DAVIDSON <S: LK C T.l-.IU' : i:i<l'i:i T ()!•■ NI'I RATIC ON WIII-:A'I'. I 53 



seem to be affected l)y the various fertilizer tre.'ilnicnts. In the work 
of Lc Clerc and his associates there was a distinct correlation 1)e- 
twecn the percentage of yellowberry, the ])rotein content, and the 
weight per i,ooo kernels. The results of this experiment, while 
establishing a definite correlation between the percentage of nitrogen 
and the yellowberry, do not indicate any such correlation between the 
protein and the weight per i,ooo kernels. It is possible that this 
difference is due to the change in the variety of wheat. In the 
former experiments hard winter wheat was used, while in this ex- 
periment soft winter wheat was used. It is, however, also possible 
that the causes affecting the protein content and color of grain and 
those affecting weight per i,ooo kernels are not the same. Attention 
is drawn to the fact that the variation in the protein content in the 
former experiments of Le Clerc and his associates is much greater 
than in the one reported here. 

Table 3. — Weight per bushel and weight per 1,000 kernels of wheat from plots 
to which nitrate of soda was applied in various quantities and at various 
stages of growth. 



Fertilizer added at 
each application. 



Stages ot growth. 



Weight per bushel. 



Water applied. 



Ferti- 
lizer in 
solu- 
tion. 



Ferti- 
lizer in 
solid 
state. 



No 
water 
applied. 



Weight per 1,000 kernels. 



Water applied. 



Ferti- 
lizer in 
solu- 
tion. 



Ferti- 
lizer in 
solid 
state. 



No 
water 
applied. 



2 lbs. NaNOs . . 

Do 

Do 

1 lb. NaNOs . . . 

Do 

Do 

2/3 lb. NaNOs . 

2 lbs. NaNOs + 
2 lbs. KCl 

Do 

Do 

2 lbs. KCl 

Do 

Do 

Check 



First 

Second 

Third 

First and second 

Second and third 

First and third 

First, second 
and third 

First 

Second 

Third 

First 

Second 

Third 



59-9 
S8.3 
S9.I 
59-9 
57-7 
59.4 
60.5 

59-4 
59-9 
60.2 

59-4 
59-4 
60.5 

59-9 
60.2 
60.2 

59-7 
60.2 
59-4 
^9.7 
59-7 
59-7 
60.5 
59.1 
59-4 
59-1 



59-4 
58.8 
59-7 
59-9 
59.7 
59-4 
60.5 

59-9 
59-7 
60.5 
58.6 

59-4 
60.5 

59-4 



57.7 
57-2 
60.2 
59-7 
60.5 
58.3 
S9-I 
59-4 
59-4 
59-9 
59-4 
59.1 
59-4 
59.4 
59-9 
59-9 
60.5 
60.2 
59-7 
59-1 
58.8 

59-9 
S9.I 
60.2 

59-7 
59-1 



29.2 
27.2 
25.1 
26.0 
25.0 
25.1 
28.4 
30.5 
27.2 
26.1 
28.0 
28.6 
30.3 
28.9 
28.0 
3I-I 
27-3 
27-5 
27.0 
27-3 
28.S 
27.8 
28.6 
27.5 
26.5 
26.3 



27.9 
27.0 
25-9 
25-7 
26.0 
24.1 
28.3 
27.1 
27.0 
27.4 
26.8 
27.2 
28.3 
28.6 



25-4 
27.8 
28.0 
26.2 
28.8 
27.2 
26.8 
26.6 
26.4 
27-3 
29.0 
28.3 
27.0 
27.1 
29.9 
28.4 
27.6 
26.7 
29.4 
26.4 
28.3 
28.1 
28.3 
28.0 
28.0 
26.9 



59-4 
58.3 



25.6 
23-7 



154 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Summary, 

1. The presence of sodium nitrate in the soil at the early stages of 
growth stimulated the vegetative grow^th of the crops and consequently 
gave greater yields. 

2. The presence of sodium nitrate in the soil at the time of heading 
gave a better quality of grain with reference to color and protein 
content. The vegetative growth was, however, not in the least 
affected. 

3. The presence of sodium nitrate in the soil at the milk stage of 
the grain had no effect on yield, quality, or protein content of the 
grain. 

4. The same results were obtained from the plots which received 
their nitrate in solution and those which received it in solid form, 
except that the yields from the plots which received the fertilizer at 
the first stage were higher in the former case than in the latter. The 
reason for this is probably the better distribution of the fertilizer 
when applied in solution. 

5. The use of potassium chloride did not affect the vegetative 
growth, nor did it appreciably affect the composition of the grain, but 
it did seem to increase the amount of yellowberry when used alone. 

Plant Chemistry Laboratory, Bureau of Chemistry, 
U. S. Department of Agriculture, 
Washington, D. C. 



« 



bailey: quality of wkstkrn si»rin(; whicat. 



155 



THE QUALITY OF WESTERN-GROWN SPRING WHEAT.^ 

C. H. Bailey. 

The crop year 1916 presented an unusual situation with regard to 
the marketing of the wheat grown in the Pacific Northwest. Ordi- 
narily this wheat, grown principally in what is known as the Inland 
Empire, is marketed through Tacoma, Seattle, and Portland. Cer- 
tain abnormal economic conditions, including high freight rates from 
the western ports and limited tonnage, have resulted in its transporta- 
tion eastward by rail. Not since the rust-damaged spring wheat 
crop of 1904 has so large a quantity entered the markets of the 
Central States. Because of the short crop of rust-damaged wheat 
produced in the spring wheat area during the crop year of 1916, the 
attention of Central States millers is again being drawn to this 
western wheat, and an unusually large quantity is being utilized by 
them. Only a portion is being ground in these mills, however, much 
of it being consigned for export through eastern ports. 

In addition to the varieties commonly grown in the Pacific North- 
west, a considerable acreage of Marquis wheat was sown this past 
season. A part of this was shipped to the Minnesota markets during 
September, 1916. Some of the first consignments were so soft and 
starchy as to render difficult a classification based on the physical 
characteristics of the kernels. The yellowberry condition had re- 
sulted in the loss of the usual angularity of the edges of the cheek 
and crease of the kernel, which had a rotundity common to certain 
other varieties. 

Tests were made of some of these first shipments of Marquis wheat, 
and the results showed them to be inferior to the average of the same 
variety of wheat grown in the Great Plains area. The writer there- 
fore proceeded to those sections of Montana, Idaho, and Washington 
which were shipping grain of this variety, and collected a number of 
samples in order to ascertain the comparative quality of each. To 
afiford a comparison with the varieties of wheat more commonly 
produced in the same areas, a number of samples of other kinds of 
wheat were collected at the same time. Th^se were tested in the 

^ Contribution from the Minnesota Grain Inspection Department Laboratory. 
Received for publication January 27, 1917. 



156 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

manner usual to this laboratory, namely, by grinding them in the 
experimental roller mill, thus producing a white middlings flour which 
was baked into bread under controlled conditions.^ The percentage 
of crude protein in the wheat was also determined. 

It was found that the greater part of the Marquis wheat produced 
in the Inland Empire is raised in the Palouse district near Pullman, 
Wash., and Moscow, Idaho, and in the Nez Perce district, particularly 
around Genesee and Lewiston, Idaho, and on the Camas Prairie be- 
tw^een Reubens and Grangeville, Idaho. This variety is particularly 
adapted to the higher altitudes of the Camas Prairie, since it matures 
sufficiently early to avoid the frost of late summer. The VoUmer- 
Clearwater Company of Lewiston, Idaho, has been largely instru- 
mental in introducing and encouraging the production of Marquis 
wheat in this portion of the Nez Perce district. 

Marquis wheat was first grown in these localities in any consider- 
able quantity during the season of 1914; the increase was sold for 
seed, and it was not until the fall of 1916 that it was shipped to any 
extent. The larger part of the Marquis wheat acreage was sown in 
the spring, although occasional fields were fall-sown. In and around 
Culdesac, Idaho, it was reported that the larger part was being sown 
in the fall ; this is the only locality visited where the reverse was not 
true. The experiment station at Moscow, Idaho, reported materially 
larger yields from the spring-sown plots of this variety than from 
those sown in the fall. 

It was observed that in general those samples of Marquis wheat 
grown at the lower altitudes were decidedly high in their content of 
yellowberry kernels, while those grown at higher altitudes were gen- 
erally corneous and dark red in color. The chemical analyses and 
baking tests of these samples paralleled their appearance quite closely, 
the corneous samples 'being higher in crude protein and in baking 
strength than those consisting largely of yellowberries. The relation 
between altitude and hardness of the grain was not exact in all cases, 
however, certain other environmental influences occasionally counter- 
acting the general relation observed. The samples collected at Lewis- 
ton, Culdesac, and Ferdinand, in Idaho, serve to illustrate the differ- 
ences between lots grown at different altitudes. Lewiston is at an 
altitude of 742 feet, although the adjacent farming land is somewhat 
higher than this ; Culdesac is 1,620 feet above sea level, and Ferdinand, 
3,728 feet. The average percentages of crude protein in the Marquis 

2 The milling methods are described in Minn. Agr. Expt. Sta. Bui. 131 and 
143 ; the baking method in Jour. Indus. Engin. Chem., 8 : 53-57, Jan., 1916. 



bailey: (juaijtv of wicstkrn spring wiikat. 



'57 



wheat s:ini])los collected at these three stations were as follows: 
lA'wiston, 8.93; Culdesac, 9.83; and Ferdinand, 11.38. The last 
average does not include the sanijile taken from a lot which had been 
allowed to become dead ripe before harvest (Lab. No. 207a), and 
which contained only 7.87 percent of crude protein. The relation 
between altitude and protein content is doubtless due to the materially 
shorter season at the higher altitudes. It is generally true that, other 
things being equal, a short growing season results in higher per- 
centages of protein and harder kernels than when the seasons are 
long. 

The Marquis wheat grown on the experiment station farm at 
Pullman, Wash., was materially higher in protein content and in 
baking strength than any of the other varieties produced there of 
which samples were obtained. The Turkey (winter) and Early 
Baart (spring) were about equal, the latter being slightly lower in 
percentage of protein, but giving a larger loaf. The yield of flour 
was greater from the Turkey, however. The common soft white and 
red wheats grown at Pullman were inferior in these respects to the 
three varieties mentioned, the Jones Winter Fife being the poorest, 
followed closely by Red Russian and Fortyfold. Two samples of 
Marquis wheat grown near Pullman and obtained through Mr. W. M. 
Chambers were decidedly inferior to the sample from the experiment 
station. They contained 8.09 and 8.49 percent, respectively, of crude 
protein and were correspondingly low in baking strength. 

The Marquis wheat obtained at the Idaho experiment station, 
Moscow, contained almost the same percentage of crude protein as 
the Pacific (Palouse) Bluestem from the same farm, but gave a larger 
and better loaf of bread. No Early Baart wheat was obtained at 
Moscow. 

Lind, Wash., in what is known as the Big Bend district, was visited, 
and a number of samples of spring wheats typical of this district 
were obtained. The Big Bend district is characterized by its low rain- 
fall, and the samples of Early Baart and Pacific Bluestem grown in 
the vicinity of Lind reflect that condition in their relatively high 
protein content and baking strength. The average percentage of crude 
protein in the Early Baart samples was 13.10, and in the Pacific 
Bluestem, 12.74. The average loaf volume was practically the same 
in the case of both varieties, being 1,473 ^^^^ i;470 c.c, respectively, 
while the average of the expansimeter tests was the same, viz., 687 c.c. 
There was a decided difference in protein content and baking strength 
between the samples of these two varieties grown in the Big Bend 



158 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

district and those grown at Pullman, Wash., in the Palouse district, 
the latter being lower in both respects. 

The larger part of the wheat grown in the Big Bend is spring sown. 
Early Baart ordinarily matures several days earlier than the Pacific 
Bluestem; this past season the fomer was affected by a drought just 
before it matured, which reduced the yield, while a late rain stimulated 
the Bluestem to a greater development, and the yields of it were 
larger than those of the Early Baart. Under normal conditions the 
property of maturing early would be an advantage, since a rain in the 
late summer is unusual in this district, and the Early Baart should 
accordingly be preferred to Bluestem. The farmers are dissatisfied 
with their experience with the Early Baart, however, and the proba- 
bilities are that less of it and more of the Bluestem will be seeded in 
the spring of 1917. 

In the principal wheat-growing sections of Montana a thaw fol- 
lowed by a freeze early in the spring of 191 6 killed most of the fall- 
sown Turkey wheat. A large part of this acreage was accordingly 
reseeded in the spring to Marquis wheat, and the total production of 
the latter was much in excess of the normal. A number of samples 
of Marquis and Turkey wheat were obtained in different parts of the 
Gallatin Valley and in the Judith Basin. The dry-farmed Marquis 
wheat of the 191 6 crop grown in the Gallatin Valley was apparently 
quite uniformly high in protein and also in baking strength. The 
sample of irrigated Marquis grown at Amsterdam, in this valley, 
was much poorer in quality than the dry-farmed wheat from a neigh- 
boring farm. A sample of 1915-crop Marquis, said to be typical of 
that crop from near Belgrade, Mont., proved to be inferior to the 
1916 crop from the same station. Many of the 1916-crop samples of 
this variety were badly mixed with Turkey winter wheat, which 
survived the freeze and matured its grain. The Turkey wheat grown 
in the Gallatin Valley, while fairly high in protein content, did not 
yield as satisfactory loaves of bread as the spring wheat from the 
same localities. 

Marquis wheat raised tributary to Lewistown, Mont., in the Judith 
Basin, also possessed very satisfactory milling and baking qualities. 
There was less difference between the Marquis and the Turkey 
samples obtained in the Judith Basin in these regards than there 
was in the Gallatin Valley samples. Such differences as were found 
were, on the average, in favor of the spring Marquis, although there 
was an overlapping in the cases of the individual samples. 

Data from the milling and baking tests and on the crude protein 
content are given in Table i. 



bailey: quality of wkstkkn si'rinc; vviikat. 



'59 



Tarle I. — Pafii obtained from viillituj and baking tests of Washington, Idaho, 
and Montana wheats, with the f^ereentagc of crude protein in each. 



WKSTKRN HARD STKlNCi WllKATS. 



Lab. 


Variety. 


Source. 


flour. 


.nsi- 
test. 


c3 S 


i-o 


i 


i 


a 


No. 



H 


We 









X 

H 


u 2 ^ 

a 








Percent. 


c.c. 


c.c. 


Pevcetit. 






PeYcent , 


167a 


No. I West- 




















ern red 


i\ .IT . JJJ.^.^ .... 


75-8 


560 


1.340 


6t 7 
u 1 . J 


08 


94 


8.21 


i68a 


IVf n rn 1 1 i<i 

i.Vi CIA Ulo 


M P "J 9 n c '7 

IN .i. . o-^vo-^ .... 


75-2 


540 


1,290 


62.6 


99 


90 


7-89 


1690 


Marquis No.i 


N.P. 44410 .... 


75-6 


650 


1,300 


62.6 


98 


96 


903 


1 70a 




CP. 120128 


75-6 


690 


1,360 


63.7 


100 


97 


9.46 


171a 


. .do 


N.P. 29785. ... 


73.6 


730 


1,440 


63-7 


lOI 


99 


II. 51 


iSoa 


Marquis 


Pullman Wash. 


73-6 


750 


1,510 


64.6 


102 


100 


11.46 


187a 


do 


do 


74-9 


690 


1.330 


65-3 


98 


98 


8.09 






189a 
i88a 


. .do 


do 


74-0 
72.2 


660 


1,360 
1,310 


67.5 
66.8 


100 


99 
98 


8.49 
8.21 


do 




680 


99 


193a 


. .do 


Moscow, Idaho 


75-4 


650 


1,390 


67.1 


100 


98 


10.06 




do 




74.0 


610 


1,310 


63.5 


lOI 


100 


8.70 


197a 

1980 


. .do 


do 


74-9 
•75-2 


660 


1.360 
1,320 


62.6 


102 


99 
94 


8.92 
9.18 


do 


do 


630 


63-3 


102 






200a 


. do 


Fort Lapwai, 






















75-1 


670 


1 .450 


62.2 


lOI 


99 


10.54 








20ia 


(Jo 


("^iilHpGQr* TH5*Hr» 


73-2 


730 


1,460 


65.5 


_ _ _ 
■103 


100 


10.49 


202a 


A/T p rn 1 1 1 « ('f a 1 n 

Marquis 


do 


72.6 
72.8 


650 
570 


1,360 
1,290 


62.2 


100 


93 
93 


9.41 
9.59 


203a 


do 


62.8 


lOI 






204a 


. .do 


Ferdinand, Ida. 


72.4 


690 


1,380 


66.2 


lOI 


98 


10.55 


205a 


do 


do 


72.4 


700 


1,430 


65.5 


102 


100 


11.69 






2060 


Marquis (cut 






















do 


69.8 


660 


1,430 


65-3 


100 


99 


II. 91 








207a 


Marquis (cut 






















do 


70.5 


500 


1,280 


64.0 


lOI 


99 


7.87 








2i8a 


No. 2 


Lewistown, 
Mont 




















70.0 


700 


1,470 


64.8 


102 


99 


1 1.46 








219a 
223a 
225a 


No. I 


do 


72.8 
70.9 


740 
750 


1,510 
1,560 


64.8 
65.5 


98 


95 
100 


12.34 
10. II 


Marquis 


do 


100 


Marquis (dry- 
farmed) . . . 


Amsterdam, 






Mont 


70.5 


760 


1.550 


65.S 


98 


99 


13.22 








226a 


Marquis 




















(irrigated) . 


do 


72.5 


600 


1,300 


63.5 


99 


98 


9-63 








227a 


Marquis 


Manhattan, 






















72.3 


740 


1. 510 


63-3 


100 


98 


12.14 


228a 


Marquis 


Belgrade, Mont. 






(1915) • • • • 




75-1 


630 


1.350 


63.1 


lOI 


100 


8.83 


229a 


Marquis 




















(1916). . . . 


do 


72.7 


680 


1,480 


67.7 


97 


97 


11.23 




Average 




73-3 


666 


1.394 


64-3 


100 


97^ 


10.01 



WESTERN HARD WINTER WHEATS. 



173a jTurkey. . . 
196a jTurkey 

j (mixed) 
199a iTurkey. . . 
217a . .do 



Pullman, Wash. 

Lewiston, Idaho 

do 

R.I. 31803 
(Spokane) . . . 



74.1 


560 


1,300 


66.2 


98 


94 


10.07 


72.3 


560 


1,400 


68.4 


99 


98 


8.21 


73.1 


670 


1,420 


66.0 


102 


99 


11-57 


72.6 


620 


1.450 


65.1 


99 


98 


10.29 



l60 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



WESTERN HARD WINTER WHEATS. — Continued. 



Lab. 
No. 


Variety. 


Source. 


Total flour. 


Expansi- 
meter test. 


Loaf 
volume. 


Water 
used. 


Color 
score. 


Texture. 


Crude 
protein in 
wheat. 


220a 

22ia 

222a 
224a 

230a 


No. 2 Mont. 

Winter. . . . 
No. 3 Mont. 

Winter. . . . 

Turkey 

. .do 

. .do . . 


Lewistown, 

do 

do 

Manhattan, 

Mont 

do 


Percent. 
70.8 

69.0 
73-6 

72.7 
70.5 


C.C. 

700 

630 
610 

610 
600 


C.C. 

1. 510 

1,490 
1.450 

I.3IO 
1.350 


Percent. 

69.9 

66.2 
67-3 

66.2 
67.7 


100 

99 
100 

98 
98 


97 

94 
95 

95 
99 


Percent. 
12.34 

12.65 
9.29 

11.26 
11.63 


Average 




72.1 


618 


1,410 


67.1 


99 


96^1 10.81 


WESTERN SOFT RED WHEATS. 


172a 
174a 

190a 
2 15a 

163a 


Wash. Hybrid 
No. 123 . . . 

Jones Winter 
Fife 

Red Russian 

. . . .do 

Jones Winter 
Fife 

Crail Fife. . . . 


Pullman, Wash. 

do 

do 

Moscow, Idaho 

Lind, Wash. . . . 
Montana 


71.6 

71.7 
72.2 
72.8 

71.9 
71.2 


560 

430 
430 
530 

480 
510 


1,250 

1,120 
1,160 
1,220 

1,120 
1,110 


67.3 

62.2 
62.2 
62.6 

61.3 
56.8 


97 

99 
96 
95 

98 
99 


95 

92 
92 
85 

75 
90 


9.30 

9.26 

8.93 
10.29 

9.12 
9.86 
9.46 


Average 





71.9 


490 


1,163 


62.1 


97 


88 


EARLY BAART WHEAT. 


i8ia 
209a 
2ioa 
2iia 
2i6a 


Early Baart . 

do 

do 


Pullman, Wash. 
Lind, Wash. . . . 
do 


71-3 
70.2 
71.6 
70.9 

69.6 


660 
730 
670 
660 

620 


1.350 
1. 510 
1.490 
1,420 

1,440 


62.0 
62.6 
62.2 
62.0 

59-3 


100 
102 
100 

lOI 

99 


99 
98 
96 
99 

98 


9.82 
13.91 
12.48 
12.92 

11.38 


. . . .do 


... .do 


. . . .do 


R. I. 35117 
(Spokane) . . . 




Average 




70.7 


668 1 1,442 


61.6 


100 1 98 


12.10 



PACIFIC BLUESTEM WHEAT. 



212a 

213a 
214a 

185a 

192a 

195c 

208a 


Pacific Blue- 
stem 

do 


Lind, Wash. . . . 
. . . .do 


70.4 

69.0 
69.6 
72.7 
73-9 
73-3 

72.2 


730 

620 
710 
510 
480 
520 

520 


1,500 

1.430 
1,480 
1,280 
1,280 
1.350 

1.300 


61.3 

61. 1 
62.4 
60.8 
63.3 
61.5 

60.2 


100 

100 
lOI 

96 

95 
97 

94 


99 

99 
98 
94 
90 
94 

92 


12.94 

12.31 
12.97 

9.21 
10.03 

8.35 

8.26 


do 


do 


do 

do 

do 

do 


Pullman, Wash. 
Moscow, Idaho 
Lewiston, Idaho 
Ferdinand, 
Idaho 








71.6 


584 


1,374 


61.5 


97I 


95 


10.58 






OTHER WESTERN WHITE 


WHEATS. 








176a 


Little Club 




















(fall) 


Pullman, Wash. 


72.7 


570 


1,290 


63.5 


94 


95 


10.74 


1 86a 


Little Club 




















(spring) . . . 


do 


73.2 


550 


1,320 


61. 1 


96 


96 


9.27 


177a 


Wash. Hybrid 






No. 128 


do 


73.5 


490 


1,210 


60.8 


95 


90 


10.20 











H.\ii.i:\': oi'AiJiA' oi- wI':sti:rn si-rinc wiii;a'I'. 



OTHER WESTEKN WHITE WHEATS. — Continued. 



Lai). 
No. 

179a 

1 S. A n 
1 o4» 

178a 
182a 
1830 
191a 


■ 1 

Variety. Source. 

i 


Total flour. 


Expansi- 
meter test. 


Loaf 
volume. 


r 

Water 
used. 


1 Color 
score. 


[ Texture. 


Crude 
protein in 
wheat. 


Wash. Hybrid 
No. 143 (tall) 
Wash. Hybrid 

No. 143 

(spring) . . . 
Fortyfold 

Dicklow 

Red Allen 
Fortyfold. . . . 


Pullman, Wash. 

do 

do 

do 

do 

Moscow, Idaho 


Percent. 
71.9 

71.9 
72.4 
71.6 
71.8 
74.1 


C.C. 
480 

520 
410 
570 
620 
460 


C.C. 
1,220 

1,280 
1,180 
1,290 
1,310 
1,170 


Percent. 
63-3 

59-1 
60.4 
61.7 
60.0 
61.7 


96 

97 
97 
98 
100 
97 


95 

97 
92 
95 
96 
70 


Percent. 
9.89 

8.61 
10.57 
8.16 
9.38 
9.72 








72,6 


SI9 


1,252 


61.3 


96§ 


92 


9.62 



Summary. 

The quantity of Marquis wheat produced in the Pacific Northwest 
and Montana during the crop year of 1916 was much larger than 
usual. This was due to an increased acreage of this variety in cer- 
tain sections of the Inland Empire, and to the reseeding of the winter 
wheat fields which had been frozen out early in the season in sections 
of Montana. 

Marquis wheat grown at Pullman, Wash., was higher in protein 
content and baking strength than any of the common varieties of 
which samples were obtained. The lots of this variety v/hich were 
grown at the lower altitudes were in general materially lower in 
baking value and percentage of crude protein than those grown at the 
higher altitudes. The difference is attributed to the shorter growing 
season under the latter conditions. 

Early Baart wheat samples which were grown in the Big Bend 
district of Washington near Lind were higher in the percentage of 
crude protein and nearly as satisfactory from the baking standpoint 
as the average of the spring wheat produced east of the divide and in 
the northern Great Plains district. 

Marquis wheat produced inider dry-farm cpnditions in Montana 
was of good milling and baking quality and was somewhat superior 
in these respects to the Turkey winter wheat grown in the same 
districts. 

The soft red and white wheats of the Inland Empire district, such 
as Jones Winter Fife, Little Club, Red Russian, and Fortyfold, are 
generally inferior in baking qualities to Marquis and Turkey wheat 
grown in the same sections. 

Minnesota Grain Inspection Department Laboratory, 
Minneapolis, Minn. 



1 62 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



GREEN MANURING: A REVIEW OF THE AMERICAN 
EXPERIMENT STATION LITERATURE— 3. 

A. J. PlETERS. 

Western Section. 

The western section includes the Great Plains, Intermountain, and 
Pacific Coast States and the Western Provinces of Canada. In this 
section the green-manure or rotation work has been conducted with a 
variety of plants rather than with any one species particularly. In 
the western United States alfalfa is the greatest leguminous forage 
plant and in these States also the moisture supply rather than the 
humus content is often the critical factor. 

CALIFORNIA. 

Wheat yields were increased on sandy soil in the San Joaquin 
Valley by turning under green manures as reported in Bulletin 211 
(1911). One plot was fallowed while on others horse beans, 
Canada field peas, rye and vetch, and rye alone were turned under in 
the spring of 1908 and the plots fallowed till the next fall, when wheat 
was sown. The average yield of wheat during 1909 and 1910 was 
as follows : 



After fallow 33-3 bushels. 

After horse beans 37-6 bushels. 

After Canada field peas 36.5 bushels. 

After rye and vetch 540 bushels. 

After rye 52.3 bushels. 

After wheat (in 1909 only) i57 bushels. 



The stand of horse beans and peas is said to have been fairly good 
and that of rye excellent. In this experiment the summer fallow 
after plowing under the green crops gave opportunity for complete 
decay and the rye gave better results than the legumes. 

On the heavier soil at Davis, Cal. (Bui. 211, p. 268), wheat after 
wheat (average of two plots) yielded 35.6 bushels; wheat after green 
manures turned under, 44.3 bushels; and wheat after fallow (one 
plot), 41.6 bushels. Here also there was a small but evident benefit 
from green manure. 



imkti:ks: (;ki:i:n manukinc 



A comparison of <]jrccn manuring with fallowincf and with con- 
tinuous croppini]^ on the yield of wheat is made in Bulletin No. 270. 
AH plots were fallowed in 1906-7, after which some received green 
manure, others were fallowed, and on two plots wheat was grown 
continuously for six years. The results show that the total yields for 
six years on the continuous grain plots was a little better than on the 
best treated plots for three years but also that on the continuous grain 
plots there was a steady and great decrease in yields while on the 
treated plots this decrease, if any, was small. A consideration of the 
yields during the last four years of the experiment shows that gen- 
erally the green-manure plots have yielded a little less than the fallow 
plot, w^hich has yielded more in one year than the continuous plot has 
in two. The author concludes that at present it is profitable to fallow 
but not to grow and turn under green-manure crops. 

In an experiment conducted at the Citrus Station, Riverside, Cal. 
(Report, 1913-14, p. 62), the relative value of a legume or a non- 
legume proved to be very dif¥erent from that in the San Joaquin ex- 
periment referred to above. The experiment at Riverside w^as com- 
menced in 1909 and each year winter green-manure crops w^ere grown, 
plowed under in spring, and corn, potatoes, and sugar beets raised in 
the summer. Barley w^as the nonlegume used and to some of the 
plots on which barley was turned under nitrate of soda varying in 
quantity from 270 to 1,080 pounds per acre was applied after the 
summer crop began growth. The report of 191 3-1 4 gives the 4-year 
average yield of corn, 2-year average yield of potatoes, and 2-year 
average yield of beets, as shown in Table 10. 



Table 10. — Yields of corn, potatoes, and beets at Riverside, Cal., after legu- 
minous and nonleguminous green-manure crops. 





Yield per acre. 


Previous treatment. 










Corn. 


Potatoes. 


Beets. 




Bushels. 


Bushels. 


Tons. 


Average on all legume plots 


40.00 


226 


17.9 


Average on barley plots without nitrate of soda (4 plots) 


29-75 


161 


12.3 


Barley plot with 270 lbs. nitrate of soda 


32.00 


166 


12.5 


Barley plot with 540 lbs. nitrate of soda 


42.00 


204 


15.7 


Barley plot with 810 lbs. nitrate of soda 


34-00 


191 


16.0 


Barlev plot with 1,080 lbs. nitrate of soda 


41.00 


218 


17.7 



On some of the legume plots, as on that on which Melilotus indica 
was turned under, the yields considerably exceeded any on the barley 
plus nitrate plots, the yield of corn being 46 bushels, potatoes 252 
bushels, and beets 19.8 tons. 



164 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

CANADA, 

At both the Brandon and Indian Head experimental farms a series 
of rotations were commenced in 1898 in which a legume was turned 
under every third year. At the Brandon Farm the old series was 
abandoned in 1905 and a new one commenced, but at the Indian Head 
Farm the records are complete for ten years. While certain minor 
changes in the rotations were made toward the end of the series the 
main plan has been carefully carried out. The record from the 
Indian Flead Farm is the more satisfactory, but in general the results 
at the two farms are in harmony.-"' 

At the Indian Head Farm 22 plots were laid out in 1899. Nos. i 
to 5, inclusive, had grain in 1899 1900 and a legume turned under 
in 1 901 ; Nos. 6 to 10, inclusive, had a legume in 1899 and grain in 

1900 and 1901 ; No. 11, rape in 1899, wheat in 1900, and fallow in 

1901 ; Nos. 12, 13, and 14, grain two years with fallow in 1901 ; Nos. 
15 and 16, continuous grain; Nos. 17 to 21, inclusive, grain in 1899, 
legume in 1900, and grain in 1901 ; while No. 22 was fallow in 1900 
and had grain in 1899 and 1901. There were thus three series of 
plots in which a legume to be turned under appeared every third year. 
The work on these plots ended with the crop of 1909. 

In Table 11 the yields of wheat for the last year in which this crop 
appeared have been given for the legume rotation plots, together with 
the yields on fallow and continuous rotation plots so far as possible. 
Since oats, barley, and rye were also raised, on some plots these crops 
occurred in the years in question and so all plots are not included in 
the table. The results shown indicate that even after three or four 
crops of legumes had been turned under in the course of ten years 
the average yield was not increased over that from fallowed land, 
though the increase over continuous grain growing was marked. 

In the report for the year ending March, 1910 (p. 352), the average 
yields of grain for the 5-year period from 1905 to 1909 on the 
Brandon Farm are given. On these plots legumes had been turned 
under one or two years with grain (mostly wheat) for three or four 
years. The fallow plots had been in grain four years and fallow 
one. The average yield of wheat only on plots that had had two 
legume crops turned under was 33.4 bushels. The average yield on 
three plots fallow one year and in grain four years was 35.12 bushels, 
while the continuous grain plots produced an average of 29.13 bushels 
of wheat. It is clear that at Brandon and Indian Head for the time 

^ Reports on this series are found in the following Experimental Farms 
Reports — 1899, 1900, 1901, 1902, 1903, 1904, 1905, 1906, 1908, 1909, 1910, and 1912. 



I'iKTKKs: (;k1':1':n manurinc 



during which this scries of experiments was conducted, green manur- 
ing has shown no benefit over fallowing. An isolated test of sweet 
clover fallow at the Brandon h\arm (Report 1898, p. 274) showed 
that wheat' yielded better after fallow than after sweet clover, but 
better after sweet clover than after wheat stubble. 



TxVRLE II. — Yields of zchcat on the Indian Head, Sask., Farm in 1907, kjoS, or 
igoQ, in s-ycar rotations including one legume, in similar rotations 
including fallow, and on continuous grain plots. 



Legume crops since 



Yields of wheat, bushels per acre. 



1907. 



2 crops of soybeans, i of alsike 

3 crops of peas 

3 crops of tares 

3 crops of red clover 

3 crops of alfalfa and alsike 

Average for legume plots 

No legumes, fallow three times 

2 crops soybeans, i of peas 

2 crops peas, i of tares 

2 crops tares, i of alsike 

3 crops of red clover 

3 crops of alfalfa and alsike 

Average of legume plots 

Average of 3 fallow plots 

Average of 2 continuous grain plots 

4 crops of peas 

4 crops of tares 

3 crops of soybeans, i of alsike 

4 crops of red clover 

4 crops of alfalfa and alsike 

Average of legume plots 

No legumes — 2 crops rape, timothy in 1905 
fallow in 1901, 1904, and 1907, wheat in 1908 
Fallow in 1901, 1904, 1907, wheat in 1908. . . 
Continuous grain 



18.17 
15.93 
12.93 
23.30 
12.40 

16.55 
23.10 



31.27 
32.00 
30.60 
29-83 
31.67 
31.07 
33.12 
14.18 



COLORADO. 

Potatoes on alfalfa sod yielded from two to five times as much 
as on land that had been manured two years before and had raised 
one crop of grain and one of corn (Bui. 57). A gain in the yield 
of potatoes on alfalfa land is also reported in Bulletin No. 117. 

KANSAS. 

In 1891 an extensive series of rotation experiments was planned 
(Bui. 20). Reports of yields from these plots were made in 1892 
(Bui. 33) and in 1894 (Bui. 47), but no conclusions can be drawn 
from them. The last reference found to this series is in Bulletin 128 
(1904), but nothing of importance is added to the previous informa- 



1 66 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

tion. It is to be regretted that a series so carefully planned was not 
carried out to its conclusion, as valuable data might have been 
obtained. 

In Bulletin 144 (1907) records are given from a 2-year rotation 
experiment in which wheat followed various crops. The average 
yield of wheat after wheat on all plots during 1905 and 1906 was 
31.9 bushels; wheat after soybeans, 28.8 bushels; and wheat after 
millet, 33.0 bushels. Soybeans was the only leguminous crop used. 
All crops were harvested. The average yields of corn for the three 
years from 1904 to 1906 on a few representative plots selected from 
the entire record (Bui. 147) were as follows: 



After oats 51.49 bushels. 

After flax 53.12 bushe'.s. 

After millet 57.21 bushels. 

After corn 5370 bushels. 

After corn and cowpea catch crop 54-30 bushels. 

After corn and rye catch crop 54-20 bushels. 

After soybeans 67.50 bushels. 

After potatoes 69.96 bushels. 



While these rotations have not run long enough to warrant the 
conclusion that cowpeas as a catch crop have no special value, the 
authors do conclude that " the second largest average yield of corn, 
67.50 bushels per acre, was produced after soybeans, and offers a 
good illustration of the value of legume crops for increasing the avail- 
able nitrogen in the soil, preparatory to growing large crops of corn 
or other heavy nitrogen feeding crops." They appear to overlook 
the fact that the yield after potatoes was larger than that after soy- 
beans and that when a legume (cowpea) was used as a catch crop 
the yield of corn was not materially increased in the time of the 
experiment. It may be noted here that in Bulletin 100 (1901) the 
statement is made that " the yields of crops of all kinds is increased 
where they follow soybeans, wheat showing in large fields an increase 
of 5 bushels per acre when following soybeans over that grown 
on adjoining land that had not been in beans. This increase is shown 
where soybeans bearing no tubercles have been grown." No experi- 
mental data for this statement are offered. 

In Bulletin 160 (1909) 5-year averages of corn and wheat follow- 
ing one year of cowpeas or soybeans are given. Wheat averaged less 
after the legume, which is said to have been due to lodging as a 
result of rank growth. Corn yields were higher after legumes, 66.53 
bushels, as compared with 52.30 bushels for continuous culture. 



riKTKRS : (;ri:kn manuuinc, 



167 



In 1906 it was found (Bui. 175. i(>i r) that wheat following cowpeas 
on newly broken sod ^ave larger yields than when following corn or 
small grain. The yields after fallow on this newly broken sod were 
about the same as after cowpeas. 

W. E. Watkins in the annual report of the Allen County Farm 
Bureau for 191 5 has presented the records for a number of i-year 
experiments on the value of sweet clover and cowpeas to precede corn 
and wheat. The various records have been condensed in Table 12. 

Table 12. — Yields of zvhcat and corn after one year of sweet clover or cowpeas, 
as compared with yields from nonlegume plots. 



Yield in bushels per acre. 



Crop. 


After 
sweet clover. 


Check. 


After 
cowpeas. 


Check. 


Wheat 


16.0 


9.0 






Corn 


54-0 


32.4 


62.5 


21.0 


Do 


35-0 


25.0 


46.6 


26.8 


Do 






40.0 


30.0 



NEBRASKA. 

A rotation and green-manure experiment was carried on from 1907 
to 1914 in which oats and other grains followed various crops (Bui. 
155). The green-manure crops, field peas and rye, were plowed 
under when in bloom and were followed by oats only. The average 
yields over the eight years show that while on spring-plowed land 
oats after oats yielded 16.3 bushels per acre, after rye plowed under 
they yielded 21.7 bushels per acre and after peas plowed under, 22.2 
bushels. After corn and after spring wheat the yields were 19.6 and 
19.4 bushels respectively, while on summer-tilled land 27.4 bushels 
were produced. While the use of green-manure crops returned a 
larger yield than continuous grain culture the yields were smaller 
than on fallow land, and the legume green-manure crop was prac- 
tically no better than rye. The authors add that the additional ex- 
pense of the green-manure crop made its use unprofitable. In the 
Annual Report of the Nebraska Corn Improvers' Association, 1912, 
pages 73-75, a member records an observation showing that under 
certain conditions the plowing under of alfalfa may have a bad effect 
on the succeeding crop. Corn failed to produce ears while on ad- 
joining land on which a legume never had been grown a crop of 40 
bushels per acre was harvested. 



1 68 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

NORTH DAKOTA. 

In 1892 an extensive rotation experiment was planned and out- 
lined in Bulletin 10. The land had been uniformly cropped from 
1883 to 1891. The author, W. M. Hays, says that ''this piece of 
land is as nearly uniform, so far as the eye can judge, as any 40 acres 
of land I ever saw." Partial reports on this work may be found in 
Bulletins 11, 23, 39, and 48, and a complete record of yields is pub- 
lished in Bulletin 100, on which the following discussion is based. It 
was inevitable that all plots should not be in wheat each year and so 
comparisons can be made only for certain years and between certain 
plots. 

Plot 19 was continuously in wheat for eight years, from 1892 to 
1899, inclusive, but 2 to 3 pounds of red clover seed was sown with the 
grain, the clover being turned under for the next crop of wheat. 
Plots 14, 24, and 25 were the nearest continuous wheat plots without 
clover. The average annual yield was nearly 2 bushels higher on 
plot 19 than that from plots 14, 24, and 25. In a letter Professor 
Shepperd has also pointed out that in the eight years there have 
been on plots 14, 24, and 25, three crops larger and twenty smaller 
than those on plot 19." The yield on plot 19 was also from 2 to 4 
bushels more than that obtained from plots 21, 22, and 23, on which 
other small grain alternated with wheat. 

On plots 7 and 8 field peas were grown one year in four ; the peas 
were cut on plot 7 and turned under on plot 8. On page 35 the 
authors analyzed the results from these plots, comparing the yields 
with those on continuous wheat plot 2, to show that a gain of more 
that 2 bushels to the acre must be attributed to green manuring, this 
being the difference between the yields of wheat on plots 7 and 8 in 
favor of the plot on which peas had been turned under. 

If the turning under of a leguminous green-manure crop under the 
conditions of this experiment was markedly beneficial it might be 
expected that the effect would be cumulative and that the later years 
of a rotation would show much better yields on such plots than on 
others not having received a leguminous green-manure crop. It will 
be interesting in this connection to compare the yields from plots 4, 7, 
8, 9, II, and 12, all of which were carried through the four courses. 
These yields are given in Table 13. 

It appears from this table that the average yield for the 16 years 
was smaller on the plot on which peas were grown or were turned 
under than on those on which millet or rape were grown. Further, 
during the last of the four courses the yield from the plots on which 



PIETICKS : CRICI'.N MANHKINC, 



l^cas were grown and tnrncd nndcr was less than that on which millet 
was grown and removed and no more than that on which ra])e was 
growni. During igo6, the sixteenth year of this experiment, and after 
four crops of peas had heen turned under on plot (S, the yield w^as 3.4 
])ushels less than on the adjoining i)lot 9, from which a crop of millet 
had heen removed every time a crop of peas was plowed under on 
plot 7. In considering the value of a legume under the conditions of 
the experiment the ahove facts must be borne in mind as well as the 
fact that the yields of wheat were larger when the peas were turned 
under than when they were removed. 



Table 13. — Average yields of wheat in bushels per aere in each of four courses 
of various rotations at the North Dakota station. 



Cropping. 


ist course, 
1892-95. 


2d course, 
1896-99. 


3d course, 
1900-03. 


4th course, 
1904-06. 


Average. 


Fallow one year, wheat 3 years . . 


19.68 


25.18 


26.92 


15-17 


21-73 


Peas I year, cut, wheat 3 years . . 


20.30 


21.78 


16.58 


10.90 


17-39 


Peas I year, turned under, wheat 












3 years 


18.35 


20.26 


22.09 


12.82 


18.34 


Alillet I year, cut, wheat 3 years. 


19.98 


23-52 


21.84 


14.55 


19.97 


Millet I year, turned under. 














20.11 


21.86 


24.03 


16.07 


20.27 


Rape I year, wheat 3 years 


23-95 


22.11 


19.78 


12.35 


19-55 



Green-manure tests have been conducted at the Edgeley, Dickinson, 
and Williston substations from 1907 to 1914, wheat and oats being 
grown in 4-year rotation of grain, corn, grain, green manure (Bui. 
no). In another rotation fallow was substituted for the green 
manure. Spring and winter rye, peas, and sweet clover were used as 
green-manure crops expect that at Williston no sweet clover was 
used. The table of yields (p. 187) shows that no effect was ap- 
parently produced by plowing under green manures ; this is also the 
conclusion expressed by the authors (p. 186). Wheat returned larger 
yields after rye turned under than after peas turned under (Third 
Annual Report, Dickinson Sub-experiment Station, 1910) and the 
following year oats on the same land also yielded better on the rye 
plot than on the pea plot. 

At the Edgeley substation wheat was grown for six years following 
various crops ; the average yields for the period are given in the Tenth 
Annual Report of the Edgeley Sub-experiment Station, 191 2. The 
highest wheat yields follow^ed corn and fallow, while the yield after 
peas and rye had been turned under was about 2 bushels less. 



I/O JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



SOUTH DAKOTA. 

Twenty-two different rotations were started in 1897 and these are 
reported in Bulletin 79, where detailed analyses of results will be 
found. The author concludes that the best average yields of wheat 
have followed corn, roots, and fallow ; that following peas and vetch 
the yields have been slightly under the general average and that the 
lowest yields followed oats or continuous wheat. In a discussion of 
each rotation the author maintains, however (page 31), that turning 
under peas has been beneficial, but the argument seems forced, and the 
author himself has doubts in regard to the validity of the evidence 
(P- 32). 

The same rotations are discussed in Bulletin 98, in which the crops 
harvested during eight years are considered. The author's conclu- 
sions may be quoted in part : 

The best average yields of wheat have been obtained in those rotations where 
that crop follows either corn or potatoes. Following these crops in the order 
of their merit as a preparation for the growth of wheat comes summer fallow, 
millet, vetch, peas, wheat and oats. 

Plowing under peas for green manure has not as yet shown any benefits over 
a summer fallow. 1 

At the end of eight years land that has grown wheat and corn alternately is 
producing better crops of wheat, of both straw and grain, than is similar land 
upon which wheat has been alternated with vetch and with summer fallow. 

A number of varieties of barley were grown in 1910 on corn 
and on turned alfalfa sod (Bui. 124). The average yield was less 
than a half bushel more on the alfalfa sod than on the corn land. 

WYOMING. 

Land that had been five years in alfalfa was broken up and wheat, 
oats, and potatoes were planted oii this and on adjoining land that had 
been tilled. Half of each plot of wheat, oats, and potatoes in 1899 
was on old alfalfa and half on tilled land. The yields were in each 
case much better on the alfalfa sod ; this was especially the case with 
oats, the yield of which was three times as large on the alfalfa sod 
as on the tilled land (Bui. No. 44). 

SUMMARY. 

In this section various legumes as well as rye and millet have been 
tried as green manure, and wheat and corn have been the chief 
indicator crops. 

With the exception of the experiments reported from California 



iMKTKRs: (;r[:i:n manukinc. 



the record shows little, if any, benefit from turninj]^ under green- 
manure crops. There are isolated cases of increased crops, as in 
Jvansas after sweet clover and cowpeas, hut the extensive tests at 
Brandon and Indian Head in Canada, in North and South Dakota, 
and in part in Kansas and Nebraska show as good or better yields 
after fallow as after a green crop turned under. This is true not only 
of the grain immediately following the green manure, but is generally 
true for the entire period of the test, which in some cases continued 
for ten years. Besides fallow, a hoed crop appeared to be a better 
preparation for grain than the turning under of a green crop or the 
stubble of a legume crop. The growing of clover and timothy for 
two years gave positive results in Kansas, the grain crop being ma- 
terially increased, especially in the later years of the rotation. 

In many cases rye and millet turned under gave as good results 
as the turning under of field peas. The turning under of an alfalfa 
sod was generally followed by larger crops than were taken from 
adjacent land on which alfalfa had not been grown, 

In California the turning under of green crops in spring resulted 
in better grain yields the following winter than were taken from 
fallowed land. In this case, however, there was really a summer 
fallow succeeding the turning under of the green manure. Here, too, 
rye gave better results than the legume. 

Under irrigation in southern California the legumes appear to have 
fully filled the place of nitrogen gatherers generally assigned them, 
since the yields following legumes were larger than those after barley, 
even when reinforced with considerable applications of nitrates. 

Addenda. 

In part i of this article, add the following under the heading 
" North Carolina" on page 72 : 

In a test conducted at the Tarboro test farm in 1901 it was found 
that the yields of corn following soy beans and velvet beans turned 
under was greater than where commercial nitrogen had been added. 
The yield after velvet beans was particularly good. (Board of Agri- 
culture, vol. 23, no. I, p. 23.) 

General Summary. 
While most of the experimental work in the South and Atlantic 
Coast sections is open to serious objection when critically considered, 
a body of evidence remains, cumulative in its effect, to show that 
leguminous green-manure crops increase the yields of following crops 
under the conditions prevailing in those sections. There is no ade- 



1/2 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

quate evidence to show whether or not this result would also be 
obtained by the use of nonleguminous green-manure crops. The 
results at the Rhode Island station point at least to the possibility 
that under some conditions the roots of a grass sod may be quite as 
efficient as a legume. Some of the Michigan and Minnesota work 
might be interpreted in the same way, though there is little evidence to 
this effect at present. At the Maryland station better yields were 
produced after rye than after crimson clover. 

In a general way, it has been shown that care must be exercised in 
the turning under of the green-manure crop to avoid injury, but the 
relation between different methods of treatment and the various fol- 
lowing crops remains to be studied further. 

While it has been shown that there is a marked residual value to a 
green-manure crop, the value of this has not been worked out. In 
fact, the subject has been touched on at but one station and that in- 
adequately. This residual value of various green-manure crops for 
the different succeeding crops in the rotation is a promising field, 
awaiting agronomic research. 

In the north there is convincing evidence of the value of red clover 
as a green-manure crop under the conditions prevailing in the Prov- 
inces of Ontario and Nova Scotia. The fact that there is a pro- 
nounced residual value has also been brought out. Farther west in 
our Ohio and Mississippi valleys evidence of the value of green 
manure crops is mostly wanting. Contrary to the Canadian results, 
there is even some evidence against concluding that it has value and 
much of the work is unsatisfactory. As a rotation crop, clover has 
been shown to be beneficial, but there is need for more exact work to 
show, if possible, how much of this benefit is due to the clover as a 
legume and how much must be ascribed to the growing of crops in a 
rotation. The question is not one of the benefit to be derived from 
a rotation ; that is well established. The important matter is whether 
a legume in such a rotation gives better results than would be obtained 
from growing a crop of similar habit but a nonlegume, as timothy. 
On this phase of the question the record throws no light. 

While in Illinois part of the evidence is negative, this becomes 
even more the case in the Northwest, where there is no adequate evi- 
dence to show that a leguminous crop is better than a nonlegume as 
green manure, nor even that turning under a green-manure crop 
results in better yields than summer fallow. It does not appear that 
any serious effort has been made to determine whether this result is 
due to the possible fact that the soils of our west and northwest 



pikti:rs: (iui:i-:N manukinc;, 



173 



still contain a sufficient supj)!) of lunnus, or to the different moisture 
conditions prevailing-. 

It is a fact of some interest to observe that, excluding; the re<(ion 
west of the Rocky Mountains, the value of a legume as green manure, 
as shown hy the published work of the American experiment stations, 
decreases roughly front southeast to northwest. There can be no 
doubt of the value of cowpeas and similar cro])s in the South, nor of 
crimson and red clover in the East and Northeast, but this feeling 
of assurance is changed to one of uncertainty as the records from the 
Ohio and Mississippi valleys are studied and finally to the conviction 
that in the Dakotas and in the Canadian Northwest the conditions do 
not warrant the use of a leguminous green-manure crop. 

Tabular Summary of Literature. 

Table 14 contains a summary of the references to station bulletins, 
circulars, and annual reports in which the effect of the preceding crop 
on the growth or yield of succeeding crops is reported. References 
to the effect of green manures on certain minor crops are given in a 
footnote to the table. 

Bibliography. 

ALABAMA. 

Newman, W. H. 

1890. Field experiments with cotton, peas, Melilotus and corn. Cane- 
brake Sta. Bui. 7. 
DuGGAR, J. F., and Richeson, J. M. 

1907. Experiments with cotton and corn in 1906. Canebrake Sta. Bui. 24. 
Stevens, F. D. 

1908. Experiments with cotton and oat's in 1907. Canebrake Sta. Bui. 25. 

1909. Fertilizer tests with cotton. Cover crops. Alfalfa yields. Cane- 

brake Sta. Bui. 26. 

1910. Agricultural value of nitrogenous materials for cotton on the Hous- 

ton clays, as determined by field trials. Residual effect of cover 
crops. Alfalfa — yields and effect as a means of restoring fertility. 
Canebrake Sta. Bui. 27. 
DuGGAR, J. F. 

1897. Experiments with oats. Ala. Agr. Expt. Sta. Bui. 95. 
1901. Corn culture. Bui. iii. 

1901. The cowpea and the velvet bean as fertilizers, Bui, 120. 

1903. Agr. Expt. Sta. i6th Ann. Rpt. 

1905. Agr. Expt. Sta. i8th Ann. Rpt. 

1905. Corn culture. Bui. 134. 

1909. Experiments with oats. Bui. 137. 

1909. Crimson clover. Bui. 147, 



174 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



o, - 



1° 



00 



00 



o ^ 



OS 

-J ^ 



(No'"' 
vO O - 



On 
to ro 



ro 00 
O M 
0\ . 

. a 

Pi d 

00 



i3 § 



o 



^ m m M 

ii s C d 



iMi':ri':Rs: (iKi-;i;N manl'rinc 



!/:> 





o 




o 


o 


0\ 


0\ 


'. T 


00 


o 































00 
. 00 



00 



^ 00 M 



03 



03 
03 

- • ;_ (U 03 ^ 

< ^ < u 

3 03 



03 

-a 

03 
C 
03 

i ^ 

o 



(U flj (L> 



o 
S 

I S - c 



1/6 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



OS -O 



00 

" d ■ 
a, • 



2- ° 



- OD 

o 









OS 








P OS 




On 
O 




00 • 

OS <N 
00 ^ 










6 


ro 


M 






^d 



OS ° 



• Os 




• 




: u 




• o* 




• ^ 




: u 


Dt. 




Pi 



Os lO U-) ^ 
Os O O (N w 

00 - M M 

^ oo" ^" °o 
_ o 
a 

5s 



o vO ^ o 



u 



S P 

OS . 

00 O 
Os" 
00 - 



















Os 


vO 00 . 




O 


0> 00 vo 








." ON 




-txO ^ 




o' 














: c- 




ro 0) ^ 


: ^ 















03 




C! 


;^ 


03 




U 






> 




o 




U 





II -g § Q .2 



03 



T3 

03 2 
03 



,1^ (f) 



c5 ^ 



OS 



riKTi'.Rs: (;ki:i:n manuuinc 



177 



0.00 



<-> XT) 
0\ M 



- Lo 

. O do <N 

O 0\ M - 

(N M wOO 



On 00 M ^, NO 
>-H (N f/5 N On 

^' M O 



8n 

^ — s - M 

00 ro 
00 M 



ON 


a 




M ^ 


IT) c5 




M VO 

. On 




On 


0" 


ONJ 


NO 



3 c d 5 ^ 



o 



CO 

<0 



-a .2 



• <D CO 

a; 

B 
W 



1/8 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



0\ 

o 


o 




Rpi 
















00 


00 






o 




0\ 








o 


M 


d\ 


o 








0\ 





0\ 
O 

00 . 

M O 



I 



0\ M 



o\ 

00 

■•-> o 

ft M 

o 

o 



(I4 



to 



6 S i^js 

C o C Cl ^ ^ 



o 



20 



oi Oj 



1 ■ 1 1 "r !•: us : ( ; k i-: i-: n iM a n u r i n ( ; . 



W9 



t 2 



w O \ 

n i OS'S 



o 

M oc 

6 ^ ^ 
2 ^ a 
a'- 
.0^ 



™ 03 



5< W ^ W 



« ^-1 CO 



u 



c; J2 M Q . ^ 



c/5 ^ U S M (/3 



oj O 



l80 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



->-> O 

a- 



o 

-J o 
in 



ft 2 - 



<< 



. M n i : "S 



O ™ 



iMirri'.KS : (;ki:i:\ .m anuring. 



IcSl 



O 00 

o o 



00 o 





O 


o 






O 


o 






0\ 






o 










oo' 






N 






a 




00 




o 


















M 






00 












a 















fS M M 







04 






o 






6 


q3 


a 


c 


a 


a 


U 


U 



M On 



- 0\ 



?^ <1J a; • 



- -Q 

Q cs 



u ^ < u w <: 

u o 

> > 



03 

< ^ < U U 

03 



5 

03 a> 



I82 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



2 I 
o 



y5 ^ W 



^ II 
II • 



Pi 



03 <-M 

U o 



CU 



u 

< :3 

in 

•• ^! 

(U w 

o 

c II 



o ^ 
o 



(L» O "-I 

3 ^ 

^ -s ^ s 

II .2 S ^ ^ 

O nS 

C ^ C ' rt Jri 

03 +J CU M-l '-^ 



P 2 

C C 03 

O CO 



^ g ^ HH O 

f> oj "r cv3 

1!^ Il o 

< 1 1^ u 2"^ 

^ j_. (J M-H to 
"(A 



o 

o ^ 
o 



o a, 



2"^ 

ON „r 



di M-l 

Pi o 



> -l-> 
O 



rt oj 'rt ^ 
^ M-i On 

QJ O (IJ 



3 ^ .(^ 



in 

CU . ^ 

M-i _<U 

o3 

a 
U 



i-( o 

II 

c3". 

>> <v 

•5 



Oh O M-i 
W 1— I 

. o < 
O " ^■ 

^ o ^ 

c/5 O 

Tl On 
M-i 5 CO 
O ^ 01 

u <-M 

oj o 

^ o" 

2 •• 

^ ?d 



3 5 

<u ^ 

M-i o 

o _ 



s < 



^ O 
O 



5 ^ 



o 

^ o 

o ^ ho 

? § 9 



o 

ct3 m 



< .s 



C CM ^ 

I °§ 

5 

<^ 

^ rvi <u 
> .. u 



be . 

C CD 

03 O 

H ON 



2 -d 13 

<U Oj Q 

03 ... 

u, o 

C CU — ' 

03 > ^ 

U O 



*^ 'a 

(U 



<u 



O «+H 

OJ CO 
CU l-l 

O 



^ > 

CU 

o 



CU 



be (J 



r[i;Ti:Ns: (;ki:i:n manuriiNc 



arkansas. 

Bennett, R. L. 

1891. Some cotton cxpcriiiuMits at Newport. lUil. 18. 
Newman, C. L. 

1891. Sorgluim and sugar-cane culture; sirup and sugar making on small 
farms; some field experiments with cantaloupes and corn. Bui. 22. 
Bennett, R. L., and Irry, G. B. 

1893. Late crops for overflow lands ; Corn : varieties for all sections of 
the state; Corn culture; Rotation of crops; Cotton: Egyptian 
varieties; Cotton culture; Stack frame for curing and storing 
cowpea hay; Cowpea hay; Forage plants; Oats for hay. Bui. 27. 

1896. Experiments with manures and rotation for improving worn cotton 
soils ; Experiments on beef and pork production in connection 
therewith. Bui. 46. 

1898. An experiment in grazing a corn and cowpea field with steers ; Ex- 

periments with peanuts, legume manuring, cotton meal, whole and 
crushed cotton seed manuring, and varieties of cotton. Bui. 58. 
Newman, C. L. 

1899. Wheat experiments. Bui. 62. 

1900. Oat experiments. Bui. 66. 

1901. Cowpea experiments. Bui. 70. 
1903. Cowpea experiments. Bui. 77. 

1901. 14th Ann. Rpt. 

california. 

Shaw, G. W. 

191 1. How to increase the yield of wheat in California. Cal. Agr. Expt. 
Sta. Bui. 211. 
1914. Agr. Expt. Sta. Ann. Rpt, 1913-14. 
Madson, B. a. 

1916. A comparison of annual cropping, biennial cropping, and green 
manures on the yield of wheat. Bui. 270. 

CANADA. 

Saunders, Wm., and Shutt, F. T. 

1902. Clover as a fertilizer. Central Expt. Farms (Ottawa) Bui. 40. 
Zavitz, C. a. 

1908. Alfalfa or luzerne. Bui. 165. 
1895. Experimental Farms Reports for 1895. (Published as appendix to the 

report of the Minister of Agriculture.) 
1896-1906. Experimental Farms Reports for 1896, 1897, 1898, 1899, 1900, 1901, 

1902, 1903, 1904, 1905, and 1906. 
1908-1912. Experimental Farms Reports for 1908, 1909, 1910, 191 1, and 1912. 

colorado. 

Cooke, W. W. 

1899. Farm notes: Alfalfa, corn, potatoes, and sugar beets. Colo. Agr. 
Expt'. Sta. Bui. 57. 
Bennett, E. R. 

1907. The Colorado potato industry. Bui. 117. 



i; 



184 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

CONNECTICUT. 

1899. Conn. (Storrs) Agr. Expt. Sta. 12th Ann. Rpt. 

1900. Conn. (Storrs) Agr. Expt. Sta. 13th Ann. Rpt. 

delaware. 

Neale, a. T. 

1891. Soil and crop test's. Agr. Expt. Sta. Bui. 11. 
1892. Del. Agr. Expt. Sta. 5th Ann. Rpt. 

GEORGIA. 

Redding, R. J. 

1894. Fertilizer, culture and variety experiments on corn ; Fertilizer and 
variety tests on cotton ; Green manuring with cowpeas. Ga, Agr. 
Expt. Sta. Bui. 27. 

ILLINOIS. 

1888. 111. Agr. Expt. Sta. ist Ann. Rpt. 
BuRRiLL, T. J., and McCluer, G. W. 

1890. Field experiments with corn. 111. Agr. Expt. Sta. Bui. 8. 
Morrow, G. E., and Hunt, T. F. 

1891. Field experiments with corn. Bui. 13. 
Morrow, G. E., et al. 

1894. Corn and oats experiments, 1893. Bui. 31. 
Gardner, F. D. 

1895. Corn experiments, 1894. Bui. 37. 
Davenport, E., and Eraser, W. I. 

1896. Corn experiments, 1895. Bui. 42. 
Hopkins, C. G. 

1903. Soil treatment for wheat in rotations, with special reference to 
southern Illinois soils. Bui. 88. 
Hopkins, C. G., and Readhimer, J. E. 

1905. Soil treatment of the lower Illinois glaciation. Bui. 99. 

1907. Soil improvement for the worn hill lands of Illinois. Buh 115. 
, and Eckhardt, W. G. 

1908. Thirty years of crop rotation on the common prairie soil of Illinois. 

Bui. 125. 
Hopkins, C. G. 

1905. Soil improvement for the Illinois corn belt. Circ. 96. 
1905. Soil treatment for wheat on the poorer lands of the Illinois wheat 
belt. Circ. 97. 
Hopkins, C. G., et al. 

1912. Hardin County soils. Soil Rpt. 3. 

1913. McDonough County soils. Soil Rpt. 7. 
, and Aumer, J. P. 

1915. Potassium from the soil. Bui. 182. 
Hopkins, C. G., et al. 

1915. Pike County soils. Soil Rpt. 11. 

1916. Winnebago County soils. Soil Rpt. 12. 
1916. Tazewell County soils. Soils Rpt. 14. ^ 



ril'/l'llKS : Ckl'.l'N M AM'RI N(; 



INDIANA. 

Latta, W. C. 

1888. Experiments with wheat; Crop rotations. Ind. Agr. Expt. Sta. 

Bui. 16. 

1888. Ind. Agr. Expt. Sta. ist Ann. Rpt. 
Latta, W. C. 

1889. Field experiments with wheat. Bui. 27. 

1890. Field experiments with wheat. Biil. 32. 
i8g2. Field experiments with corn. Bui. 39. 
1893. Field experiments with corn. Bui. 43. 

1893. Field experiments with wheat'. Bui. 45. 

1894. Field experiments with corn and oats. Bui. 50. 
■ , and Ives, G. R. 

1894. Field experiments with wheat. Bui. 51. 

1894. Ind. Agr. Expt. Sta. 7th Ann. Rpt. 
Latta, W. C, and Ives, G. R. 

1895. Experiments with corn and oats. Bui. 55. 

1895. Ind. Agr. Expt. Sta. ^th Ann. Rpt. 

1896. Ind. Agr. Expt. Sta. 9th Ann. Rpt. 
Latta, W. C, and Anderson, W. B. 

1897. Field experiments with corn, oats, and forage plants. Bui. 64. 
Latta, W. C., and Skinner, J. H. 

1901. Systems of cropping with and without fertilization. Bui. 88. 
WiANCKO, A. T., and Fisher, M. L. 

1906. Winter wheat. Bui. 114. 

IOWA, 

Brown, P. E. 

1912. Bacteriological studies of field soils. II. The effects of continuous 
cropping and various rotations. Iowa Agr, Expt. Sta. Research 
Bui. 6. 

1915. Bacterial activities and crop production. Research Bui. 25. 
Stevenson, W. H., et al. 

1916. Maintaining fertility in the Wisconsin drift soil area of Iowa. Bui. 

161. 

Stevenson, W. H., and Brown, P. E. 

1916. Rotation and manure experiments on the Wisconsin soil area. Bui. 
167. 

KANSAS. 

Georgeson, C. C., et al. 

1891. Experiments with wheat. Kans. Agr. Expt. Sta. Bui. 20. 

1892. Experiments with wheat. Bui. 33. 
1894. Experiments with wheat. Bui. 47. 

CoTTRELL, H. M., et al. 

1901. Soy beans in Kansas in 1900. Bui. 100. 
1904. Experiments at Fort Hays Branch Station, 1902-1904. Bui, 128. 
TenEyck, a. M., and Shoesmith, V. M. 

1907. Small grain crops. Bui. 144. 
1907. Indian corn. Bui, 147, 

TenEyck, A. M., and Call, L. E. 
1909. Cowpeas, Bui, 160. 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



TenEyck, a. M. 

1911, Grasses. Bui. 175. 
Watkins, W. E. 

1916. Ann. Rpt. of the Allen Co. Farrai Bureau for the year 1915. Pub- 
lished at lola, Kans. 

MARYLAND. 

Miller, R. H., and Brinkley, E. H. 

1895. Potato experiments. Md. Agr. Expt. Sta. Bui. 31. 

1896. Potato experiments. Bui. 38. 

1897. Corn and potato experiments. Bui. 46. 
Patterson, H. J. 

1899. Fertilizer experiments with different sources of phosphoric acid. 
Bui. 68. 

1907. Results of experiments on the liming 6f soils. Bui. no. 

1907. Fertilizer experiments with different sources of phosphoric acid. 

Bui. 114. 

MASSACHUSETTS. 

KUHN, J. 

1894. Green manuring. Mass. Agr. Exp. Sta. Bui. (not numbered). 

(Translated by E. W. Allen and published as a special bulletin of 
Mass. Sta.) 

1897. Agr. Expt. Sta. (Hatch) 9th Ann. Rpt. 
1900. Agr. Expt. Sta. (Hatch) 13th Ann. Rpt. 
1903. Agr. Expt. Sta. (Hatch) i6th Ann. Rpt. 

• MICHIGAN. 

1910-1911. Mich. Agr. Expt. Sta. 24th Ann. Rpt. 

minnesota. 

Hays, W. M. 

1895. Grain and forage crops. Minn. Agr. Expt. Sta. Bui. 40. 

1895. Agr. Expt. Sta. Ann. Rpt, 1895. 

1896. Agr. Expt. Sta. Ann. Rpt., 1896. 
Snyder, H, 

1897. Effects of the rotation of crops upon the humus content and the fer- 
tility of soils; production of humus from manure. Bui. 53. 
, and Hummel, J. A. 

1904. Soil investigations. Bui. 89. 
Hays, W. M., et al. 

1908. The rotation of crops. Bui. 109. 
Walker, G. W. 

1912. The relation of different systems of crop rotations to humus and 

associated plant food. Bui. 128. 

mississippi. 

Ferris, E. B. 

1907. Report of work at McNeill branch experiment station for 1906. 
Agr. Expt. Sta. Bui. loi. 



I'lKTKKs: (;ki:kn manuring. 



187 



MISSOURI, 

Miller, M. F., and Hutchison, C. H. 

1909. Soil experiments on the upland loam of southeast Missouri. Mo. 
Agr. Expt. Sta. Bui. 83. 
Miller, M. F. 

1909. Soil experiments on the prairie silt loam of southeast Missouri. 
Bui. 84. 

Miller, M. F,, and Hudelson, R. R. 

1914. Soil investigations; Jasper County experiment field. Bui. 119. 
Miller, M. F., Hudelson, R. R., and Hutchison, C. B. 

1915. Soil experiments on the level prairies of northeast Missouri. Bui. 

126. 

Miller, M. F., Hutchison, C. B., and Hudelson, R. R. 

1915. Soil experiments on the dark prairies of central and northeast Mis- 
souri. Bui. 127. 

1915. Soil experiments on the rolling glacial land of north Missouri. Bui. 
128. 

1915. Soil experiments 'on the red limestone upland of southeast Missouri. 
Bui. 129. 

1915. Soil experiments on the gray prairie of southwest Missouri. Bui. 
130. 

MuMFORD, F. B., et al. 

1915. Work and progress of the agricultural experiment station for the 

year ending June 30, 1914. Bui. 131. 

NEBRASKA, 

• Snyder, W. P., and Osborn, W. M. , 

1916. Rotations and tillage methods in western Nebraska. Nebr. Agr. 

Expt. Sta. Bui. 155. 
1912. Nebr, Corn Improvers' Asso. Ann. Rpt., 1912. 

new jersey. 

1894. N. J. Agr. Expt. Sta. isth Ann. Rpt. 
LiPMAN, J. G., et al. 

1912. Miscellaneous vegetation experiments. N, J. Agr, Expt. Sta. Bui. 
250. 

LiPMAN, J. G. 

1912. The associative growth of legumes and non-legumes. Bui. 253. 

1912. Agr. Expt. Sta. 33d Ann. Rpt. 

1913. Agr. Expt. Sta. 34th Ann. Rpt. 

LiPMAN, J. G. 

. 1914. Nitrogen utilization in field and cylinder experiments. Bui. 281. 

1914. Agr. Expt. Sta. 35th Ann. Rpt. 
LiPMAN, J. G., and Blair, A. W. 

1916. Investigations relative to the use of nitrogenous plant-foods, 1898- 
1912. Bui. 288. 
, , et al. 

1916. Cylinder experiments relative to the utilization and accumulation of 
nitrogen. Bui. 289. 



1 88 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



NEW YORK. 

Lyon, T. L., and Bizzell, J. A. 

191 1, A heretofore unnoted benefit from the growth of legumes. N. Y. 

Agr. Expt. Sta. (Cornell) Bui. 294. 

1914. Experiments concerning the top-dressing of timothy and alfalfa. 

Bui. 339. 

NORTH CAROLINA. 

Chamberlain, J. R. 

1889. The value of pea-vine manuring for wheat. N. C. Agr. Expt. Sta. 

Bui. 72. 

1890. Value of pea-vine manuring for wheat. Bui. 77. 
KiLGORE, B. W. 

1902. Fertilizer and other tests with corn and cotton. State Bd. Agr. 
Rpt, vol. 23, no. I, p. 23. 

north DAKOTA. 

Hays, W. M. 

1893. Grain and forage crops. N. D. Agr. Expt. Sta. Bui. 10. 

1893. Grain and forage crops. Bui. 11. 
Shepperd, J. H, 

1896. Grain and forage crops. Bui. 23. 
, and TenEyck, A. M. 

1899. Crop report for 1898. Bui. 39. 

1901. Wheat farming experiments and soil moisture studies. Bui. 48. 
1910. Agr. Expt. Sta. (Dickinson Substation) 3d Ann. Rpt. 
Shepperd, J. H., and Doneghue, R. C. 

1912. Cropping systems for wheat production. Bui. 100. 
1912. Agr. Expt. Sta. (Edgeley Substation) loth Ann. Rpt. 
Thysell,' J. C, et al. 

1915. Dry farming investigations in western North Dakota. Bui. no. 

OHIO. 

Hickman, J. F. 

1892. Field experiments with wheat. Ohio Agr. Expt. Sta. Bui. 42. 
Thorne, C. E., and Hickman, J. F, 

1893. Field experiments with commercial fertilizers. Bui. 49. 
1893. Field experiments with commercial fertilizers. Bui. 53. 

Thorne, C. E. 

1905. The maintenance of fertility. Circ. 40. 
Snyder, A. H., and Cook, C. D. 

1906. The maintenance of fertility: fertility on Strongsville soil. Bui. 168. 
Thorne, C. E. 

1907. The maintenance of fertility. Bui. 182. 

1908. How to determine the fertilizer requirements of Ohio soils. Circ. 79. 
1912. Maintenance of soil fertility; annual summary tables of experiments at 

Wooster. Circ. 120. 
1914. Clermont County experimental farm : 3d Ann. Rpt. Bui. 275. 
1914. Maintenance of soil fertility: plans and summary tables of experiments 

at Wooster and Strongsville. Circ. 144. 



riF/ncKS : (;kI':i:n manuuinc, 



Williams, C. G., and Wklton, F. A. 

1915. Corn experiments. Bill. 282. 
Ames, J. W., and Boltz, G. E. 

191 5. Tobacco. Bui. 285. 

RHODE ISLAND. 

1803. R. I. Agr. Expt. Sta. 6th Ann. Rpt. 
1897. R. I. Agr. Expt. Sta. loth Ann. Rpt. 
Wheeler, H. J., and Tillinghast, J. A. 

1900. A three-year rotation of crops, potatoes, rye, and clover. R. I. Agr. 
Expt. Sta. Bui. 74. 

1900. A four-year rotation of crops, Indian corn, potatoes, rye, clover. 

Bui. 75. 

1901. A five-year rotation of crops, Indian corn, potatoes, rye, grass, grass. 

Bui. 76. 

Card, F. W. 

1903. Bush fruits Bui. 91. 
Adams, G. E., and Wheeler, H. J. 

1906. Continuous corn culture. Bui. 113. 
Hartwell, Burt L., and Damon, S. C. 

1916. A twenty-year comparison of different rotations of corn, potatoes, 

rye, and grass. Bui. 167. 

SOUTH CAROLINA. 

Newman, J. S. 

1901. Corn. S. C. Agr. Expt. Sta. Bui. 61. 

south dakota. 

Chilcott, E. C. 

1903. Crop rotation for South Dakota. S. D. Agr. Expt. Sta. Bui. 79. 
Cole, J. S. 

1906. Crop rotation. Bui. 98. 
Willis, C, and Champlin, M. 

191 1. Progress of grain investigation. Bui. 124. 

tennessee. 

Mooers, C. a. 

1911. Fertility experiments in a rotation of cowpeas and wheat. Part I. 

The utilization of various phosphates. Tenn. Agr. Expt. Sta. Bui. 90. 

1913. Fertility experiments in a rotation of cowpeas and wheat. Part II. 

The effect of liming on the crop production. Part III. The effect 
of liming and of green manuring on the soil content of nitrogen 
and humus. Bui. 96. 

1914. The rational improvement of Highland rim soils; conclusions from 

six years of field experiment's with various crops. Bui. 102. 

, and Robert, S. A. 

1914. Fertility and crop experiments at the West Tennessee Station. Bui. 
109. 



190 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



VIRGINIA. 

Ellett, W. B., Hill, H. H, and Harris, W. G. 

1915. The effect of association of legumes and non-legumes. Va. Agr. 

Expt. Sta. Tech. Bui. i. 

WISCONSIN. 

Delwiche, E. J. 

1907. Report of the northern sub-stations for 1906. Wis. Agr. Expt. Sta. 
Bui. 147. 

WYOMING. 

BUFFUM, B. C. 

1900. Alfalfa as a fertilizer. Wyo. Agr. Expt. Sta. Bui. 44. 
Evans, Morgan W. 

1916. Some effects of legumes on associated nonlegumes. In Jour. Amer. 

Soc. Agron., v. 8, no. 6. November-December, 1916. 

miscellaneous. 

Hopkins, C. G. 

1915. Soil fertility and permanent agriculture. Latest edition. 
LiPMAN, Jacob G. 

1910. A method for the study of soil fertility problems. In Jour. Agr. 

Science, v. 3 (1908-10), p. 297. 
1913. A further discussion of certain methods used in the study of " The 

associative growth of legumes and nonlegumes." In Jour. Amer. 

Soc. Agron., v. 5, no. 2. April-June, 1913. 
Lyon, T. L., and Bizzell, J, A. 

1913. A discussion of certain methods used in the study of "The asso- 

ciative growth of legumes and nonlegumes." In Jour. Amer. Soc. 
Agron., v. 5, no. 2. April-June, 1913. 
Shutt, F. T. 

1907. Chemistry and Canadian agriculture. In Science, Aug. 30, 1907. 
Westgate, J. M., and Oakley, R. A. 

1914. Percentage of protein in nonlegumes and legumes when grown alone 

and in association in field mixtures. In Jour. Amer. Soc. Agron., 
v. 6, nos. 4-5. July-October, 1914. 



Journal of the American Society of Agronomy. 



Plate 3. 







^^^^ 

■ — 1^ 


z 






1 ^ 




I 



Method of making capsules to prevent cross pollination: (i) Equipment 
for punching holes in capsule; (2) rolling gummed strip to form tube; (3) 
(lipping the capsule in the wax bath; (4) completed capsule tied to bamboo 
stick, and lath to which stick is fastened. 



\\ .\i.i.i:k AM) Til A rcii i:k : i-ki:\'icnti nc. acci-:ss oi' roLU.x. I'^i 



IMPROVED TECHNIQUE IN PREVENTING ACCESS OF STRAY 

POLLENJ 

Adolph Waller and L. E. Thatcher. 

Previous Resp:arch. 

The difficulties that have vexed the plant breeder tryin^^ to protect 
a pedigreed culture from stray pollen are as old, one might well say, 
as the recognition of the significance of pollen itself by Camerarius 
in 1694. Darwin wrote about his fears that small insects might get 
through the nets he was usiiig to protect his flowers. Recently Pearl 
and Surface- built cages around bean plants to protect them from 
bumble bees carrying pollen, while Shaw,^ on the other hand, made 
use of thrips (Thysanoptera sp.) to insure cross pollination. It 
would be a difficult and profitless task to assemble the references con- 
cerning means of guarding against adventitious pollen, for although 
everyone is aware of the possibility of its access it is only rarely that 
one finds mention of measures taken to prevent chance crossing. In- 
formation on methods of pollination is scarce also, though papers on 
corn by Collins* and Gernert^ were published several years ago. If 
descriptions of all the time-saving methods in pollination and all the 
devices used to protect the ehte plants could be gathered so as to be 
accessible it would make interesting and instructive reading It is to 
be hoped that the files of the Journal of the American Society of 
Agronomy will become the disseminator of as much of this informa- 
tion as is related to our important crop plants. 

In Hillman's^ excellent compilation of the nature and scope of the 
plant-breeding work in Germany, photographs are given of a number 

1 Contribution from the Farm Crops Laboratory, Ohio State University. 
Received for publication January 26, 1917. 

2 Pearl, R., and Surface, F. M. Studies in bean breeding. Maine Agr. Expt. 
Sta. Bui. 239. 1915. 

3 Shaw, H. B. Thrips as pollinators of beet flowers. U. S. Dept. Agr. Bui. 
104. 1914. 

* Collins, G. N. Improvements in technique of corn breeding. In Proc. 
Amer. Breeders' Assoc., v. 7-8, p. 349-352. 1912. 

^ Gernert, W. B. Methods in the artificial pollenation of corn. In Proc. 
Amer. Breeders' Assoc., v. 7-8, p. 353-367. 1912. 

^ Hillman, Paul. Die Deutsche landwirtschaftliche Pflanzenzucht. Berlin, 1910. 



192 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

of different methods employed to prevent promiscuous cross-pollina- 
tion in many kinds of plants. Muslin cages surrounding plots, caps 
and bags covering individual plants, and glass tubes placed over the 
heads of selected wheat plants are some of the devices pictured. It 
was the illustration showing the glass tubes that suggested another 
means of preventing accidental crossing in the small grains. The 
waxed paper capsule was chosen not as a mere substitute but as an 
improvement upon the tubes destined for wider uses than with the 
small grains only. 

Materials and Methods. 

During the spring and summer of 1916, the writers used the method 
described below with such success that they feel warranted in offer- 
ing it at this time to others in crop improvement work, although 
slight refinements and changes maybe made during the next few years. 
Since these are no more than anyone employed in similar investiga- 
tions is likely to try out for himself, little hesitancy is felt in narrating 
what has been learned to date. It is sincerely hoped that other ex- 
perimenters will give the suggestion contained here a thorough trial, 
returning such criticism and observations as are derived from their 
results. 

The ideal device for protecting the floral organs or an inflorescence 
from adventitious pollen would have the following qualifications : 
It must effectually exclude stray pollen ; the size and shape must be 
readily adaptable to a variety of growth habits of different plants ; 
it. must be sufficiently substantial to withstand the destructive action 
of the weather, during at least the blooming and often during the 
ripening period of the plant ; to permit the escape of excess moisture 
from within, it must be freely ventilated ; it must be impervious to 
water from the outside ; it should guard against accumulation of heat 
from the sun's rays ; the least possible weight is a desideratum, but in 
addition an easy means of fastening the protective device to a support 
ought to be at hand ; it should be convenient, so that without loss of 
time it may be applied or removed. In short, while affording pro- 
tection, it should offer as little interference as possible to the normal 
plant processes. Along with all the desirable features, it must be 
inexpensive. 

The tube or capsule which has been found to fill most nearly these 
specifications is a simple affair made from white ledger paper of 
fairly firm texture, sheets 17 by 22 inches in size and running 28 sheets 
to the pound. No elaborate equipment or special skill is required 
for its manufacture. The size which can be used successfully on the 



WAIJJ'.K AND TirA'lCI I I:K : I'KI'AI.N'II X'C ACCI'.SS Ol' l'( )l ,1 .i; N . 1(^3 

snuill grains is iiiadc by culling the pajjcr into strips 12 by 30 centi- 
meters. These are then placed several sheets thick on a block of soft 
wood standing with the grain end up. A row of holes 7 millimeters 
in diameter and 2 centimeters apart is ])unchcd across the sheets 4 
centimeters from one end by means of an ordinary leather punch, 
as shown in Plate 3, fig. i. A strip 15 mm. wide along one side of a 
sheet is then gummed and the paper rolled on a cylinder of wood about 
35 millimeters in diameter. This forms a paper tube, such as is 
shown in Plate 3, fig. 2. The lube is then slipped along the wood 
form, punched end first, until about i centimeter extends beyond the 
form. The edge is then crimped over, the end gummed, and a disk 
of gummed paper slightly larger than the tube is applied, forming a 
cap. Having made sure that the adhesion is perfect, the form is 
withdrawn. The capsule is then ready to receive the bath of wax 
made as described below. The wax is most conveniently handled in 
a tall glass cylinder the diameter of which is only slightly larger than 
the paper capsule, as shown in Plate 3, fig. 3. It may be kept in 
liquid form by immersing the cylinder in a water bath. The capsule 
is dipped about half its length, withdrawn slowly, reversed, and when 
cool, dipped the rest of the way. 

Strong wrapping twine that has been waxed in the same mixture is 
cut into convenient lengths, about 50 cm., and one of these lengths 
is tied about 6 cm. from each end of the capsule. These weather- 
proofed strings serve to fasten the capsule securely to its support. A 
loose wad of absorbent cotton is pushed up into the capsule until just 
below the ventilating holes, making the capsule ready for use in the 
field. 

The support for the capsule is best made as follows : An ordinary 
plastering lath is sharpened at one end. To the other end a piece 
of slender bamboo not over i centimeter in diameter at the base is 
fastened with small staples. The support is driven firmly into the 
ground 'by pounding on the end of the lath to which the bamboo is 
fastened. The fiat lath gives added stability, as it is not likely to be 
twisted and lifted by the wind; as would be the case with a more 
nearly round or square support. It is firm but flexible and not easily 
broken or loosened by storms. The length of the support will of 
necessity be adjusted to the height of the plant. It is well to have 
the bamboo somewhat taller than the stem of the plant at the time of 
encapsuling to allow for some elongation up to the time of ripening. 
The mounted capsule is shown in Plate 3, fig. 4. 

Having prepared the inflorescence, a capsule should be slipped 
over it and tied to the support. A loose plug of cotton is carefully 



194 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



inserted in the lower end to exclude pollen that might be blown up 
into the capsule or brought in by insects. In tying the capsule to the 
support, the twine should be wrapped around the bamboo stick at 
least twice to prevent slipping and tied in a bow knot. The proper 
tag or label can be inclosed within the capsule. Usually it is advisable 
to fasten the stem or branch to the support in two places by means 
of a waxed cord tied rather loosely. It is an easy matter to remove 
the capsule at any time for examination or for any plant-breeding 
operation. After fertilization has taken place and stray pollen is no 
longer dangerous, the cotton plugs may be removed. The capsule 
then affords excellent protection against damage by birds, storms, etc., 
and is conspicuous enough to be found easily in the field at any time. 
Grain which has ripened in these capsules is of prime quality. 

The most satisfactory preparation for waxing the capsules is com- 
posed of the following: 

Pure white beeswax 800 grams. 

Stearic acid ' 200 grams. 

This wax gives body to the capsule, making it quite firm and resilient, 
yet is not softened materially by the heat of the sun, as its melting 
point is 67° C. Paraffin with a melting point of 57° C. has not given 
good results. 

The capsules have been used with much satisfaction in selfing wheat 
(in which, in the authors' opinion, more crossing normally takes 
place than is generally supposed), rye, timothy, orchard grass, meadow 
foxtail, and Italian rye grass, and in hybridizing wheat, oats, barley, 
alfalfa, and soybeans. The usefulness of the capsules is not limited 
to the plants mentioned in the present discussion. A change in the 
size and shape of the forms upon which the tubes are rolled and in 
the means of supporting the capsules in position will easily adapt them 
for use on shrubs and small trees as well as on a wide range of 
herbaceous plants. 

It sometimes becomes necessary to replace the cotton in the capsules 
after a hard rain, as water may gain access through the ventilating 
holes and force them out. A method of construction to overcome this 
is now under consideration. 

Summary. 

The advantages of these reinforced capsules over other devices for 
guarding against undesirable cross pollination may be summarized 
as follows: 



\valm:r and rii a I ( II i:u : i'ki:vi:nti nc, a(*ci:ss of I'orj.i-.N. 195 

(ihiss tubes of the same general form as the capsules herein de- 
scribed are heavy, costly, fraj^ile, and collect moisture on the inside 
which will frequently spoil the pollen and prevent fertilization or 
favor later the growth of molds. When exposed to the direct rays 
of the sun, the temperature within the glass tubes becomes danger- 
ously high. The translucent paper capsules hinder l^ut little the con- 
tinuance of the normal plant processes and are cheap, light, and 
durable. 

Paper bags, wisps of cotton, and tissue paper coverings are not 
substantial. After the slightest precipitation or even after heavy dews 
they are likely to be soggy and require replacement. 

No other methods known to the writers will protect against the 
pollen thrips or other small insects that sometimes render pedigreed 
cultures worthless. At the same time some of the moisture given off 
from the enclosed portion of the plant is absorbed hy the cotton. 

The isolation of the inflorescences in the paper capsules obviates 
the necessity of scattering plats of cultures in places that are fre- 
quently unfavorable for the growth of plants and are also incon- 
venient to visit and care for. The capsule likewise makes possible 
the isolation of flowers borne on shrubs and trees that could not be 
separated a sufficient distance from each other to insure freedom 
from undesirable pollen. 

The large degree of protection which the capsule affords to the 
developing and ripened fruit is also a distinct advantage. 

Ohio State University, 
Columbus, Ohio. 



196 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



AGRONOMIC AFFAIRS. 

NEW BOOKS. 

The Small Grains. By Mark Alfred Carleton, Cerealist, Bureau 
of Plant. Industry, U. S. Department of Agriculture. The Mac- 
millan Company (New York), 1916. 8vo. Pages 699; figs. 183; 
bibliography of more than 500 titles. 

The material in this book, which is intended for collegiate use, is 
presented under four main divisions : — I, Cereal Plants ; II, Cereal 
Environment ; III, Cereal Crops ; and IV, Buckwheat and Rice. 

The book opens with a discussion of the form and structure of the 
small grains, with original figures showing the anatomical structure of 
the wheat seed, culm, and leaf blade. A discussion of plant growth 
and nutrition follows. The author says that a study of these chapters 
may be omitted by those students who have had good training in 
plant physiology. 

The origin, classification, varieties, and regional adaptation of wheat 
are then taken up. A fairly complete list is given of important United 
States varieties of wheat, with an extensive list of world varieties. 
These are listed under the eight commonly recognized agronomic 
groups. The United States and Canada are then subdivided into ten 
wheat districts and the important varieties for each district are given. 
Statements of the needs of these various districts as to improved 
sorts are of special value. A short discussion of the wheat distribu- 
tion in other countries includes a statement of the particular value of 
the varietal groups grown in these various regions. Oats, barley and 
rye are discussed in essentially the same manner as has been outHned 
for wheat. 

Introduction, selection, and hybridization are discussed under the 
heading of " Improvement of the Small Grains." A description of 
the original introduction and subsequent improvement of many of our 
important cereal varieties is an interesting feature. Improvement by 
selection is presented under the headings of (i) sorting and roguing, 
(2) mass selection, and (3) pure Hne selection. The essential differ- 
ences and relative value of these three methods of work are discussed. 
A more detailed presentation of the use of the seed plat would add 
materially to this chapter. The selection of enough seed for a seed 
plat of I to 2 acres would be quite a task. The selection of a con- 



ACRONOMrC AI'I'AIKS. 



197 



sidorablc luinibor of typical heads at harvest would probably accom- 
plish the same result. Neither does there seem to be auy ])articular 
value in the plan of growing- enough seed in the special seed plat for 
the entire acreage. 

The two methods of pure-line selection given by the author are the 
centgener and row plans. These methods as now practiced differ 
only in the minor particular of shape of plot. A description of the 
rod-row method now so widely used in preliminary breeding and 
varietal tests would be of value. 

Under Hybridization " the author describes the application of 
Mendel's law to inheritance and subsequent improvement in small 
grains. The behavior of the first, second, and third generations of a 
cross is given when the parents dififer by a single contrasted character. 
The author says, " When the parents dififer in two or more characters 
there is effected a recombination of characters of much value in breed- 
ing." An illustration of the behavior of crosses dififering in two or 
more characters would make the matter much clearer. A brief de- 
scription of the general class of results obtained in crosses between 
types which dififer in important quantitative characters as height of 
culm, size of seed, and yield would be of interest even though these 
characters do not give as clear-cut segregation as is often obtained for 
qualitative characters. 

The author describes a considerable number of valuable economic 
varieties which have been produced by the hybridization method. 

Part II discusses cereal environment under the headings, Soil Rela- 
tions, Climatic Relations, and Cereal Adaptation and Association. 
The important cereal regions of Europe and Asia are considered from' 
the standpoint of the nature of the soil. This is followed by a similar 
presentation of the soil of American cereal growing regions. This 
presentation considers mechanical and chemical constituents of the 
soil of the cereal regions, particular attention being paid to a discus- 
sion of moisture relations and alkali resistance. 

A summarized statement is given of the general climatic features 
common to all cereal-growing regions, followed by a more detailed 
discussion of the dififerences between the dififerent important areas. 
Especial emphasis is placed upon moisture, sunlight, and temperature. 

Under " Adaptation " the important natural cereal regions of the 
world and the main groups specially adapted to these regions are 
discussed. Acclimatization, change of seed, and effects of environ- 
ment are presented under the heading of " Environment." A discus- 
sion of the cereal crops from the ecological standpoint is given under 
the heading of The Cereal Plant Community. 



198 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Instead of discussing methods of culture and handling of the crops 
in connection with the initial treatment of each crop, the author deals 
with this question for all crops in a separate section of his book. 
This would seem to be a commendable departure. As there are few 
cultural details which apply exclusively to any one of the small grain 
crops, much repetition is avoided by this method of treatment. This 
discussion appears in Part III of the book. The subject matter is 
presented under the headings, " Soil Treatment," " Growing the 
Crops," and " Gathering the Crop." Very full discussion is given and 
experimental results are generously quoted. 

The greater portion of the discussion on uses of cereals naturally 
is devoted to wheat flour. This discussion includes a very good de- 
scription of the milling process, and a sketch showing the various 
steps in the process is presented. Grades of flour are described with 
reference both to the step in the milling process from which they are 
obtained and the kind of wheat used in their production. The edible 
pastes made from wheat are briefly discussed. The uses of cereals 
as grain and forage for live stock, and their uses in the preparation 
of breakfast foods and in malting and distilling are also dealt with. 

Under the heading, " Cereals in Commerce," the author discusses 
various phases of grain marketing. The facihties for receiving the 
grain locally and at the terminal markets are described, with a dis- 
cussion of grain grading and inspection. A detailed description of 
each of the market grades within each of the commercial classes of 
grain might well have been included. 

A few brief but pointed details regarding exchange operation are 
very appropriately given. The knowledge of such operations as 
hedging and selling for future delivery is not generally possessed \)y 
those who are not large dealers in grain or who otherwise come into 
contact with the operations of the large grain exchanges. The lack 
of such knowledge is responsible for much unjustifiable suspicion of 
the operations of the organized grain trade. 

Buckwheat and rice are taken up separately as Part IV, as these 
crops are botanically different from the four important cereals previ- 
ously discussed. The same general plan of procedure is used as has 
been already outlined for wheat, barley, oats, and rye. 

The book is written in a very clear and interesting manner by a 
recognized authority in this particular field. It is not only suitable 
for instructional work in colleges but for the most part is presented 
so clearly that it deserves a place in every cereal farmer's Hbrary. 

H. K. Hayes, 
P. J. Olson. 



ACKONOMIC vXFl'AlUS. 



'99 



MEMBERSHIP CHANGES. 

Tho nRMnhcrsliip of the Society as reported in the ATarch issue was 
64J. Since that time JJ new nienil)ers have heen added and i nienil)er 
lias resi<^ned, niakin^i>- a net gain of 21 and a total liienihership at this 
time of 663. The names and ad(h-esses of the new memhers, toj^cthcr 
with the name of the meml)er resii^ned and such chanj^^cs of address 
as have come to the notice of the Secretary, as are follows: 

New Mkmhkrs. 

Anderson. A. C, Forest Service, Ogden, Utah. 

Andrews, Myron E., Warner Dist. Agr. School, Warner, Okla. 

Baker, O. E., Farm Management, U. S. Dept. Agr., Washington, D. C. 

BuGBY, M. O., Canfield, Ohio. 

Cocke, R. P., Williamsburg, Va. 

Cramer, W. F., Station A, Ames, Iowa. 

DE Werff, H. a.. Agricultural Building, Urbana, 111. 

DouGALL, Robert, Macdonald' College, Quebec, Canada. 

Frank, W. L., 116 E. Eleventh Ave., Columbus, Ohio. 

Halverson, W. v., Soils Office, Agr. Expt. Sta., Ames, Iowa. 

Hoke, Roy, 318 West St., Stillwater, Okla. 

Jackson, J. W., Substation No. 9, Pecos, Texas. 

Kenworthy, Chester, Warner Dist. Agr. School, Warner, Okla. 

KiLLOUGH, D. T., Substation No. '5, Temple, Texas. 

Kraft, J. H., State Teachers' College, Greeley, Colo. 

Langenbeck, Karl, 1625 Hobart St., N. W., Washington, D. C. 

Letteer, C. R., San Antonio Expt. Farm, San Antonio, Tex. 

Mortimer, George B., College of Agriculture, Madison, Wis. 

Nevin, L. B., 2828 Webster St., Berkeley, Cal. 

Smith, Howard C, Bur. Soils, U. S. Dept. Agr., Washington, D. C. 
SouTHWORTH, W., Manitoba Agr. College, Winnipeg, Man., Canada. 
Wilkins, F. S., Farm Crops Dept., Iowa State College, Ames, Iowa. 

Member Resigned. 
Geo. W. Graves. 

Changes of Address. 

Bartlett, H. H., 335 Packard St., Ann Arbor, Mich. 
Bliss, S. W., Agr. Expt. Sta., Wooster, Ohio. 
Brandon, Joseph F., Akron Field Station, Akron, Colo. 
Clark, Chas. F., Box 747, Greeley, Colo. 
Craig, C. E., Otwell, Ind. 

Emerson, Paul, Maryland State College, College Park, Md. 
Jensen, L. N., Box 1214, Amarillo, Tex. 
Lyness, W. E., Cheyenne Field Station, Archer, Wyo. 
Pridmore, J. C, 616 Rhodes Building, Atlanta, Ga. 



200 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



NOTES AND NEWS. 

A. B. Beaumont, assistant professor of soil technology at Cornell 
University, has been appointed associate professor of agronomy and 
acting head of the agronomy department at the Massachusetts College. 

J. A. Foord, head of the division of agriculture at the Massachusetts 
College, is on a year's leave of absence and is pursuing graduate study 
at Cornell University. 

Arthur Goss, for fourteen years director of the Purdue University 
station, has resigned to give his entire attention to his farming interests 
near Vincennes, Ind. 

The American Association of Agricultural College Editors will meet 
at Cornell University, June 28 and 29, 191 7. 

The third Interstate Cereal Conference will be held in Kansas City, 
Mo., June 12-14, 1917, in cooperation with the Kansas Agricultural 
Experiment Station. Station and college workers, grain dealers, 
millers, and others interested in cereal production and utilization, par- 
ticularly in Kansas, Missouri, and nearby states, are invited to 
attend. One of the principal topics for discussion will be the recently 
announced Federal grades for wheat. Particulars regarding the pro- 
gram may be obtained from the secretary, Chas. E. Chambliss, U. S. 
Department of Agriculture, Washington, D. C. 



Journal of the American Society of Agronomy. 



Plate 4 




Fig. I. Some of the variations in quantities of grain from ears of the same 
length. Ears 120 and 250 had the same circumference and the same number 
of rows, but yielded 192 and 137 grams, respectively. Ears 330 and 17 had 
the same circumference and same number of rows, Init yielded 215 and 156 
grams of grain, respectiveh'. The differences in weight are due largely to 
differences in depth of kernel. 




Fic. 2. Some of the extreme variations in quantity of grain from ear sections 

of the same length. 



JOURNAL 

OF THE 

American Society of Agronomy 



Vol. 9. May, 1917. No. 5. 



THE RELATION OF COB TO OTHER EAR CHARACTERS 

IN CORN.i 

A. JE. Grantham. 
Introduction. 

One of the objects of t-he corn breeder is to produce a type of ear 
that will carry the maximum quantity of grain. This has led to a 
critical examination of the physical characters of the ear, with the 
result that many varieties of corn have been improved to such an 
extent that they exceed the percentage of grain required by the 
standard, viz., 80 percent for the commercial grades. The increase 
in the yield of grain per ear has been accomplished largely by the 
selection of seed ears with well-filled butts and tips and with deeper 
kernels of a more desirable shape. The question then arises, " Is the 
cob, as the carrier of the grain, an important character to be con- 
sidered in relation to type of ear?" To what extent the characters 
of the cob are correlated with those of the ear as a whole has not 
been carefully determined. While some biometrical work has been 
done upon the corn plant, little attention has been given to the statis- 
tical relations of the characters that determine the grain weight of 
an ear. 

It is the object of this paper to present some of the correlations 
that exist between the cob and other ear characters. The cob itself 
has been largely overlooked from a mathematical standpoint in corn 
breeding. This paper attempts to answer some of the questions 
which may arise concerning the size, weight, and density of cob in 

1 Contribution from the Department of Agronomy, Delaware Agricultural 
Experiment Station. Presented at the ninth annual meeting of the American 
Society of Agronomy, November 13, 1916. 

201 



202 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

relation to yield of grain and to type of kernel, including depth, 
thickness, and weight. In other words, is the cob of an ear of corn 
correlated with other desirable characters to such an extent as to 
warrant careful examination of this part of the ear in selecting seed? 

Methods. 

The work reported in this paper was conducted at the Delaware 
Agricultural Experiment Station from 1910 to 191 5. The data given 
were obtained in the progress of an investigation to determine the 
relation between the physical characters of ears to the vigor and yield 
of the plant. In order to get ears for seed that had a given char- 
acter well developed it was necessary to examine a large number of 
ears with respect to the number of rows of kernels, weight of ear, 
depth, thickness, and weight of kernels, and the yield of grain. 
Owing to the great irregularities and variations in the formation of 
the butts and tips of ears, it was necessary to take a section of definite 
length from each ear. This was done by cutting a section 12 
centimeters long from the middle portion of the ear, just far enough 
from the butt to eliminate all irregular kernels. Thus, all of the ears 
studied were constant with respect to length. By this means the 
relation of size, shape, and weight of kernel to the other characters 
could be established in a comparable manner. 

In these studies, data were taken from 3,500 ears. Each fall 
several hundred ears were brought to the heated laboratory where 
they remained for several weeks until both grain and cob had thor- 
oughly dried. Special attention was given to the dryness of the cob 
at time of cutting. Tests were made to determine the constancy of 
the dry weight. As the conformation of the ear is governed largely 
by the type and regularity of the kernel, care was taken to select for 
cutting and measurement only those ears which approached the cylin- 
drical in shape and carried straight rows of kernels. The Johnson 
County White was the variety studied, and it was not difficult to 
obtain ears of excellent type as to shape and straightness of row. No 
effort was made to select ears of a given type for cutting; on the 
contrary, as wide a variation of characters as possible was sought 
except in shape of ear and straightness of row. Sections were not 
made from ears under 8 inches in length, owing to the difficulty of 
obtaining a fully developed cutting. No limit was set upon the varia- 
tions of the other characters. Some of the extreme variations in cer- 
tain characters are shown in Plate 4. 



GRANTHAM I EAR C ll ARACTKRS IN CORN. 



203 



The following- data were recorded for each section : 

Weight of section. 

Number of rows. 

Circumference of ear. 

Thickness of kernel. 

Weight of shelled grain. 

Weight of cob. 

Circumference of cob. 

Depth of kernel. 

Weight of the individual kernel. 

Percentage of cob. 

Density of cob. 

The weights were recorded in grams and measurements of length 
and circumference in centimeters. The thickness of kernel was 
determined by counting the number of kernels (in situ) in 10 centi- 
meters. The average of three readings was taken. The depth of 
kernel was deduced from the difference in the diameters of cob and 
ear. The weight of the individual kernel is expressed by the number 
required to weigh 10 grams. The density of cob is given as a con- 
stant derived from the formula, Weig ht ^ While this does not 

Circumference^ 

give the actual density, it furnishes a comparable expression. 

Review of Literature. 

While some statistical work has been done upon correlations in 
corn, the greater part has dealt with the plant as a whole. Other 
workers have pointed out certain correlations, but the relationships 
have not been quantitatively determined. Very few statistical studies 
on the correlation of characters in ear corn have come to the attention 
of the writer. Apparently no biometrical studies have been made 
upon the cob in its relation to other characters of the ear. Mention 
will be made of the published work on this subject, not that it has 
direct bearing on the studies of the author but to indicate the manner 
of approach to the general subject of correlations in corn. 

Brigham (1896),^ who made an extensive study of Longfellow 
flint corn, concluded that an increase in the weight of corn is accom- 
panied by an increase in the number of kernels, weight of cob, and 
weight of individual kernels. These correlations were not established 
by statistical methods. 

Davenport (1897) found a correlation coefficient of 0.87 + 0.005 
between weight of ears in ounces and length of ear in inches, in the 

2 References are to names and dates at the end of the paper. 



204 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Learning variety of corn. He showed that there was a relationship 
between the circumference and length of ear in inches of 0.47 + 0.02. 

Thiel (1899) noted that the greater the diameter of the ear the 
larger the percentage of cob. 

DeVries (1901) found that the size of the kernel decreased as the 
number of rows per ear increased. 

Fruwirth (1904) has called attention to certain correlations he 
found in Szekler maize. He states that a larger yield of grain is 
accompanied by a larger number of kernels and by greater weight per 
individual kernel. It does not appear that these correlations were 
studied by statistical methods. 

Craig (1908) made a number of observations on the correlation 
of physical characters of the corn plant but did not calculate the 
variation constants from his data. This, together with the fact that 
he used only 50 to 100 individuals in each case, renders his results 
unsatisfactory so far as an exact degree of correlation is concerned. 

Ewing (1901) made biometrical studies on the relation of weight 
of grain to the diameter of the stalk, the length of leaf, the breadth 
of leaf, the height of the seedlings, the number of internodes, the 
length of ear at the appearance of silks, the date of the appearance 
of tassel, the date of appearance of pollen, date of appearance of 
silks, the duration of the flowering period, and the number of 
branches in the tassel. The highest correlation found was 
0.393 + 0.020, between the weight of grain and the diameter of 
the stalk. He holds that most of the correlations noted may be 
classed as variations in the fluctuating variability of the characters 
concerned and on further consideration one would put most of them 
in the class of environmental correlations." 

Correlations, 
i. circumference of cob and weight of grain. 

It is often held by corn breeders that large cobs should be avoided 
in selecting seed corn. However, it is obvious that a cob with large 
circumference will carry more grain than a small one, provided the 
character of the kernel is the same in both cases. The practice of 
selecting seed ears with cobs of moderate size is probably due to the 
fact that large ears cure more slowly. 

In Table i are given the results obtained in studying the relation 
between the circumference of cob and yield of grain with 3,500 ears 
of Johnson Co. White corn. The extremes of circumference range 
from 7 to 15 cm., with a mean of 10.5 cm.; of weight of grain, 105 



GRANTHAM : EAR CHARACTERS IN CORN. 



205 



to 285 grams, with a mean of 196. It will be seen that there is a 
positive correlation of 4148 between the circumference of the cob 
and the weight of grain, with a probable error of + .0095. As was 
mentioned above, no cars under 8 inches in length were sectioned and 



Table i. — Relation betzvcen ivcight of shelled grain per section in grams, sub- 
ject; and circumference of cob in centimeters, relative, (r = .41 18 + .00947.) 

Circumference of cob in centimeters. 


















10 










10 





10 


























d 









(N 






ro 






10 












0\ 


o\ 











































































1 otals 






















h? 






















06 


06 


On 




6 


6 










ro 






•4 




io5~i 10 














2 




















2 


1 10— 1 1 5 












I 






















I 


I I5~I20 




































120-125 










I 


2 








I 














4 


125-130 








2 


I 






2 


















5 


130-135 










I 




2 


3 





















135-140 








2 


3 


4 




3 








I 










T A 
10 


140-145 


I 




I 




I 








3 


^ 














13 


145-150 








3 


A 
U 


7 


7 




















30 


150-155 






4 




7 


g 


g 


g 


4 




3 












45 


155-160 






4 


3 


1 


20 


15 


13 


7 


3 


I 












76 


160—165 






2 


1 1 


T A 
iO 


2 2 


T A 

iO 


9 


A 



4 


3 












90 


165-170 




I 




9 


36 


35 


29 


16 


10 


8 


2 




I 








148 


170-175 






2 


13 


35 


4» 


29 


20 


13 


7 


3 












177 


175-180 




I 


3 


1 2 


30 


04 


43 


30 


23 


13 


2 












230 


180—185 






I 


7 


34 


45 


52 


49 


2o 


17 


7 












241 


185-190 








7 


38 


53 


62 


39 


40 


19 


5 












264 


190-195 






I 


4 


36 


54 


55 


68 


40 


26 


6 


2 










292 


195-200 






I 


8 


27 


50 


77 


51 


60 


18 


II 


4 


3 








310 


200—205 








6 


27 


55 


71 


79 


44 


33 


17 


3 


I 


I 






337 


205-210 








3 


14 


35 


60 


58 


43 


29 


12 


4 










258 


210-215 








I 


14 


28 


48 


53 


38 


23 


II 


8 


I 






I 


226 


215-220 








I 


9 


14 


34 


47 


48 


31 


17 


2 


2 


3 






208 


220-225 








I 


4 


II 


18 


28 


26 


31 


18 


8 


2 




I 




148 


225-230 










3 


8 


15 


34 


34 


31 


14 


6 


2 








147 


230-235 










I 


4 


10 


15 


17 


7 


8 


5 


I 


I 






69 


235-240 










I 


2 


5 


II 


6 


II 


8 


2 


3 








49 


240-245 












5 


3 


3 


10 


7 


5 


5 


I 








39 


245-250 














4 


I 


3 


5 


6 


4 


2 


I 


I 




27 


250-255 














2 


I 


3 


3 




2 










II 


255-260 












I 


I 


3 


6 


I 


2 


2 


I 


2 




I 


20 


260-265 




















I 




2 


I 




I 




5 


265-270 
























I 










I 


270-275 
















2 


















2 


275-280 
























2 










2 


280-285 
























I 










I 


Totals 


I 


2 


19 


95 355 577 672 668 513 331 161 


72 


21 


8 


3 


2 


3.500 



measured. There was no Hmit to the circumference of the ears 
selected. The data indicate that the ears with large cobs carry the 
most shelled grain. The opinion that large cobs carry shallow 
kernels resulting in a low yield per ear is not borne out by this study. 



206 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



2. CIRCUMFERENCE OF COB AND WEIGHT OF KERNEL. 

In general, cobs of large circumference have a larger number of 
rows of grain than those of small circumference. The larger the 
number of rows on an ear, the circumference being the same, the 
narrower are the kernels. Narrow kernels are likely to be relatively 

Table 2. — Correlation between zveight of individual kernels (number in 10 
grams), subject; and circumference of cob in centimeters, relative, 
(r — — .0185 + .01140.) 



Circumference of cob in centimeters. 



CO 





10 


q 


10 




II. 


10 


12.0 


10 

M 




ro 


10 









UO 















00^0 


2 





-S 


4-" 





+-> 







Q 
4-> 


Totals 







10 







IT) 









10 





10 


q 


10 










00 


°° ^ On 2 












00 






•n^ 




18 








2 






I 






I 








4 


19 








323 


2 








I 


I 








12 


20 








388 


13 




8 


3 












43 


21 








II 10 15 


14 


8 


6 


2 


I 








I 


68 


22 








4 6 24 28 


45 


27 


20 


4 


3 










161 


23 




I 




7 23 30 45 


43 


42 


21 


1 1 


3 


3 








229 


24 


I 




I 


6 31 47 48 


59 


47 


29 


13 


8 


6 








296 


25 




I 


4 


10 30 60 74 


58 


34 


31 


13 


6 


I 


3 






325 


26 






2 


14 35 68 70 


72 


49 


37 


15 


6 


I 


I 


I 




371 


27 






2 


8 36 56 75 


59 


61 


34 


13 


7 


2 


2 


I 




356 


28 






2 


8 30 47 59 


45 


47 


24 


13 


9 


2 


I 


I 




288 


29 








8 25 34 61 


=;8 


41 


20 


13 


6 




I 


I 




277 


30 






I 


4 36 46 38 


40 


38 


12 


14 


5 










234 


31 






2 


6 22 33 26 


38 


30 


18 


7 


6 










188 


32 






2 


5 21 27 31 


37 


25 


20 


8 


I 


2 








179 











2 10 20 22 


21 


15 


17 


II 












118 


' 34 






2 


7 9 16 19 


14 


12 


6 


8 


2 










95 


35 








412 10 


10 


9 


7 


4 


2 


I 








. 59 


36 








2 9 7 10 


9 


10 


I 


4 


2 










54 


37 






I 


1575 


7 


2 


6 


2 


I 










37 


38 








1285 


4 


6 


2 




I 










29 


39 








5 4 


1 1 


2 


I 


3 




I 








27 


40 








I 24 


2 


I 


I 














II 


41 








I 2 I 6 


2 


3 






I 










16 


42 








I 3 4 


I 


I 
















10 


43 








2 


2 


I 






I 










6 


44 








I 


I 


















2 


45 












I 
















I 


46 








I 




















I 


47 






























48 






























49 










I 


















I 


50 






























51 






























52 








I 




I 
















2 


Totals 


I 


2 


19 


95 355 577 672 668 513 331 161 


72 


21 


8 


4 


I 


3.500 



lighter in weight than broad kernels. On the other hand, an ear 
with a small number of rows is likely to have rather wide kernels. 
Table 2 shows that the correlation between the circumference of the 



GRANTHAM: EAR CllAKACTHRS IN CORN. 



207 



col) and the weight of the kernel is — .0185 4: .01 14, which is 
practically negligible in value. In this variety of corn the weight of 
the kernel does not seem related to the circumference of the coIj. 
The undersized kernels which are often seen on ears of small diam- 
eter are probably offset in weight by the longer kernels which are 
common in ears of large diameter. The density of cobs and kernels 
respectively is also a factor that affects the relation. It is interesting 
to note that the coefficient of variability is high for both characters, 
being 19.97 for the circumference of the cobs and 16.18 for the 
weight of the kernel. 

3. CIRCUMFERENCE OF COB AND DEPTH OF KERNEL, 

It is commonly supposed that ears with cobs of large circumference 
have shallow kernels. Table 3 shows that the coefficient of correla- 



Table 3. — Correlation between depth of kernel in centimeters, subject; and cir- 
cumference of cob in centimeters, relative, (r = — .1789 + .01104.) 

Circumference of cob in centimeters. 





10 



06 


to 
06 




6\ 


10 
d 


q 
d 


10 






K-5 





(N 





to 
ro 





to 

'^t 



10 




















































Totals 







10 





ir> 














10 









to 





to 








i> 


00 


00 

















(N 






ro 








.7- .8 














I 




















I 


.8- .9 












I 


2 


4 


I 


2 


2 


3 










15 


.9-1.0 








I 


3 


4 


6 


17 


9 


14 


8 


2 


I 








65 


I.O-I.I 






2 


6 


23 


44 


61 


64 


62 


45 


27 


17 


3 


2 


2 




358 


1.1-1.2 


I 




2 


23 


63 133 161 162 153 


95 


49 


17 


9 


3 






871 


1. 2-1. 3 




I 


7 


25 104 172 187 191 133 


92 


41 


14 


5 


I 


2 


I 


976 


1. 3-1. 4 




I 


4 


25 


97 129 151 159 


99 


51 


26 


II 


2 








755 


1. 4-1. 5 






3 


13 


54 


71 


81 


54 


44 


28 


7 


7 


I 


2 






365 


1. 5-1. 6 






I 




10 


19 


18 


16 


10 


4 


I 












79 


1.6-1.7 








2 


I 


4 


3 


I 


I 






I 










13 


1. 7-1. 8 














I 




I 
















2 


Totals 


I 


2 


19 


95 355 577 672 668 513 331 161 


72 


21 


8 


4 


I 


3.500 



tion between the circumference of cob and the depth of kernel is 
— .1789 -[- .0710, indicating that there is a moderate tendency for 
large cobs to carry shallow kernels. The extremes of depth of kernel 
range from 0.7 to 1.9 cm. with a mean of 1.2 cm. The coefficient 
of variability in kernels is 10.78, much less than that for circumfer- 
ence of cob. 



4. CIRCUMFERENCE OF COB AND THICKNESS OF KERNEL. 

An examination of Table 4 shows that there is a slight negative 
correlation between the circumference of the cob and the thickness 



2o8 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



of kernel. The correlation factor is — .1053 + .0113. This indi- 
cates that the smaller cobs are more likely to have thick kernels. 
The extremes of thickness of kernel range from 18 to 32 kernels to 
cover 19 cm., with a coefficient of variability of 8.1 1. The coefficient 
of variability is less for thickness of kernel than for any other char- 
acter studied. 

Table 4. — Correlation between thickness of kernel (number in 10 centimeters) , 
subject; and circumference of cob in centimeters, relative. 
= — .1053 + .01127.) 



Circumference of cob in centimeters. 









10 









0\ 


0\ 

1 





d 


to 






10 




CN 


10 

M 




ro 


10 

fO 




rl- 


10 



to 










1 




00 
10 


00 

1 




i-S 





M 

1 


q 


M 
U-5 


q 


to 


q 


10 




to 


Totals 












06 




d 


d 


d 


d 








oi 


CO 


ro 




4 








18 












I 


I 




I 
















3 






19 












3 




I 


2 


I 














7 







20 








I 


I 


4 


6 


5 


7 


3 


3 


2 


2 




I 




35 


_c 




21 








3 


9 


17 


18 


12 


II 


13 


6 


3 


5 


I 






98 






22 




I 


3 


2 


25 


30 


48 


69 


34 


35 


14 


3 


I 








265 


% 

1-1 


w 
u 


23 


I 






9 


37 


78 


89. 


88 


72 


61 


21 


13 


2 


I 


I 


I 


474 


ken 


<u 
a; 


24 






I 


19 


75 106 122 133 104 


72 


38 


15 


3 


2 


I 




691 


a 


25 






6 


22 


74 115 133 126 


98 


56 


28 


13 


3 


4 






678 






26 






6 


16 


47 


97 no 117 


83 


47 


28 


14 


3 








568 


u 

OJ 


(U 




27 




I 


2 


II 


44 


58 


82 


69 


47 


19 


12 


4 






I 




350 


a 




28 






I 


7 


22 


36 


34 


30 


35 


18 


6 


5 


I 








195 






29 








3 


13 


21 


13 


15 


13 


4 


4 












86 






30 
31 
32 








I 
I 


4 
3 
I 


8 
3 


13 
3 


3 


6 


2 


I 




I 








39 
10 
I 






Totals 


I 


2 


19 


95 355 577 672 668 513 331 161 


72 


21 


8 


4 


I 


3.500 



5. WEIGHT OF COB AND WEIGHT OF GRAIN. 

Table 5 shows the relation between the weight of cob and the 
weight of grain. The extremes for weight of cob range from 15 to 
75 grams; for weight of grain, 105 to 285 grams. The correlation 
factor is .3064 + .0103, indicating a fairly close relationship between 
these two characters. A comparison with Table i shows a stronger 
correlation between the circumference of cob and weight of grain 
than between the latter and weight of cob. This indicates that the 
weight of cob is a more variable factor than the circumference. The 
coefficient of variability for circumference is 9.24 + .075, and for 
weight of cob, 19.97 + .168. 

6. WEIGHT OF COB AND WEIGHT OF KERNEL. 

The weight of kernel as expressed by the number required to weigh 
10 grams ranges from 18 to 32. The coefficient of variability is 



C.RANTIIAM : EAR CHARACTERS IN CORN. 



209 



16.18 + •1344- J^l^^' factor of correlation between these two char- 
acters is — .1837 + -01 10. This shows there is no correlation between 
high percentage of cob and high weight of kernel. It predicts a 
moderate tendency for heavy cobs to produce light kernels. In 'ral)le 



Table 5. — Relation between weight of shelled grain per section in grams, sub- 
ject; and weight of cob per section in grams, (r = .3064 + .0103.) 

Weight of cob in grams. 








10 




















to 





10 










fO 


ro 






10 


10 






1 




Totals 




10 





to 










u-i 










10 







M 


CM 


CM 




ro 






10 


to 


-0 


•0 






io5~i 10 




I 




I 




















I I o~ 115 


I 


























I 15— 120 




























120-125 


I 


I 


I 




I 
















4 


125—130 


I 
























r 


130-135 


























5 


135—140 


I 




4 


5 


4 
















16 


140—145 




3 




3 


4 
















T 1 


145—150 




4 


7 


5 


5 


5 
















150-155 


I 


4 


1 2 


1 1 


1 1 


5 




I 










45 


155—1^0 




4 


20 


27 


14 


10 














'76 


160—165 




1 1 


2 1 


20 


23 


10 


5 












90 


165— 1 70 


I 


12 


29 


54 


33 


15 


4 












T /I 8 

140 


170-175 


I 


14 


34 


59 


35 


2 1 


7 













177 


175—180 




15 


39 


82 


52 


28 


10 












230 


1 8 0— 185 


I 


1 1 


42 


00 


At 


44 


13 


3 










241 


185-190 




6 


38 


80 


75 


47 


15 


3 










264 


190-195 




7 


41 


93 


70 


55 


24 












292 


195-200 




8 


50 


77 


91 


59 


21 


3 


I 








310 


2 QQ 20^ 




g 


30 


93 


90 


61 


39 


5 


4 








"J "7 

33 / 


205-210 


2 


5 


30 


53 


80 


44 


31 


10 


2 


I 






258 


210-215 




5 


21 


60 


63 


50 


16 


7 


3 


I 






226 


215-220 






21 


39 


61 


43 


25 


12 


4 


2 


I 




208 


220-225 




2 


18 


37 


40 


27 


10 


9 


5 








148 


225-230 






12 


22 


38 


43 


14 


14 


I 


3 






147 


230-235 






3 


16 


19 


20 


7 


4 










69 


235-240 




I 


3 


7 


13 


7 


13 


3 




I 




I 


49 


240-245 






3 


9 


7 


II 


7 


I 




I 






39 


245-250 






I 


4 


7 


6 


4 


4 


I 








27 


250-255 






I 


5 


3 




2 












II 


255-260 








3 


6 


4 


I 




4 


2 






20 


260-265 












2 


I 


2 










5 


265-270 












I 














I 


270-275 














I 


I 










2 


275-280 














I 


I 










2 


280-285 
















I 










I 


Totals 


10 


126 


491 


934 


909 


620 


274 


94 


29 


II 


1 


I 


3.500 



2 it was found that there was no correlation between the circumfer- 
ence of the cob' and the weight of kernel. This would lead to the 
conclusion that no very high correlation exists between the circum- 
ference of cob and the weight of cob. Large cobs may lack normal 
density and for that reason are not proportionally heavier than 
small cobs. 



2IO JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Table 6. — Correlation between weight of individual kernel (number in lo 
grams), subject; and weight of cob in grams, relative, (r = — .1837 + .011.) 



Weight of cob in grams. 








10 





10 





10 





10 





to 





LO 






M 




ro 




"vf- 




LO 


10 










Totals 




10 





to 





LO 





10 


1 




1 

LO 


1 




1 

LO 








M 




ro 


ro 






LO 


10 










18 










I 


I 


2 












4 


19 






4 


I 


3 


I 




3 










12 


20 




I 


3 


5 


16 


9 


7 


I 


I 








43 


21 




2 


10 


10 


22 


14 


6 


3 


I 








68 


22 




3 


10 


31 


42 


40 


25 


6 


4 








161 


23 


I 


7 


24 


58 


48 


56 


19 


13 


2 


I 






229 


24 




10 


26 


79 


77 


56 


27 


7 


II 


2 




I 


296 


25 


I 


1 1 


31 


70 


100 


68 


30 


9 


I 


4 






325 


26 


I 


6 


47 


118 


88 


71 


27 


II 


2 








371 


27 




II 


43 


92 


lOI 


74 


22 


7 


4 


I 


I 




356 


. 28 


I 


9 


47 


72 


80 


41 


23 


13 


I 


I 






288 


1 29 


I 


II 


35 


89 


71 


38 


23 


9 










277 


2 30 




16 


39 


78 


44 


35 


19 


3 










234 







5 


36 


52 


52 


25 


13 


4 


I 








188 


M 32 




10 


30 


59 


43 


22 


II 


I 


I 


2 






179 


.S 33 




4 


30 


24 


28 


24 


7 


I 










118 


i£ 34 


2 


6 


20 


24 


21 


16 


3 


3 










95 


£ 35 




2 


14 


13 


19 


7 


4 












59 






5 


8 


15 


17 


8 


I 












54 


^ 37 




I 


10 


1 1 


1 1 


4 














37 


Z 38 




3 


10 


9 


3 


3 


I 












29 


^ 39 




I 


2 


6 


II 


4 


3 












27 


B 40 




I 


3 


5 




I 


I 












II 


1 4> 


I 


I 


3 


4 


5 


2 














16 


42 


I 




5 


I 


3 
















10 


43 


I 




I 


2 


2 
















6 


44 








2 


















2 


45 








I 


















I 


46 








I 


















I 


47 




























48 




























49 








I 


















I 


50 




























51 




























52 








I 


I 
















2 


Totals 


10 


126 


491 


934 


909 


620 


274 


94 


29 


II 


I 


I 


3.500 



7. WEIGHT OF COB AND DEPTH OF KERNEL. 

Table 7 shows that correlation between the weight of cob and the 
depth of kernel is — .0747, with a probable error of + .01 13. The 
correlation is negative and very low. Large cobs by weight are not 
accompanied by deep kernels. Large cobs by volume, as shown in 
Table 3, tend to carry shallow kernels. 

8. WEIGHT OF COB AND THICKNESS OF KERNEL. 

An examination of Table 8 shows a correlation of — .15 between 
the weight of cob and the thickness of kernel, with a probable erro 
of +-OIII- There is moderate indication that heavy cobs tend t 



GRANTHAM : EAR CHARACTERS IN ( ORN. 



21 I 



Tahle 7 —Correlation between depth of kernel in eeiitinieters, subject; and 
weight of cob in grams, relative. (r = — .0747 + .01 13.) 

Weight of cob in grams. 





to 





?5 
lO 


ro 

k 


»o 


i 


in 


<o 
?n 




JO 




1 










Totals 


.7- .8 




I 






















I 


.8- .9 


I 




5 


I 


3 


2 


2 


I 










15 


.9-1.0 


2 


3 


5 


1 1 


22 


15 


7 












65 


I.O-I.I 


I 


9 


38 


99 


78 


70 


50 


12 


I 








358 


1.1-1.2 


I 


28 


124 


228 


206 


166 


82 


22 


8 


4 


I 


I 


871 


1. 2-1.3 


3 


42 


114 


236 


280 


193 


73 


23 


II 


I 






976 


1.3-1.4 


I 


29 


123 


229 


182 


117 


42 


25 


5 


2 






755 


1.4-1.5 


I 


II 


66 


102 


108 


48 


14 


8 


3 


4 






365 


1. 5-1.6 




3 


12 


24 


26 


8 


3 


2 


I 








79 


1. 6-1. 7 






3 


4 


3 


I 


I 


I 










13 


1. 7-1. 8 






I 




I 
















2 


Totals 


10 


126 


491 


934 


909 


620 


274 


94 


29 


II 


I 


I 


3.500 



■5 



carry thin kernels rather than thick. In Table 4 it was shown that 
large cobs by volume were more likely to be accompanied by thin 
kernels. Small cobs both by volume and weight seem to be associated 
with thick kernels. 



Table 8. — Correlation between thickness of kernel (number in 10 centimeters), 
subject; and weight of cob in grams, relative, (r = — .1500 + .0111.) 

Weight of cob in grams. 







10 

M 


10 

cs 






ro 
10 

CN 


10 

ro 


ro 



10 

ro 


10 





10 

-si- 


10 
10 



to 




to 
to 


to 
-0 



-0 



i> 

to 


10 
i>- 


t- 


Totals 


18 




I 




2 


















3 


19 






2 


I 


2 


I 


I 












7 


20 




I 


4 


5 


8 


6 


7 


2 


2 








35 


21 




4 


10 


28 


20 


18 


12 


4 


2 








98 


S 22 


I 


8 


26 


69 


63 


62 


23 


8 


4 


I 






265 


i; 23 




9 


47 


121 


120 


98 


49 


23 


5 


I 




I 


474 


a 24 


3 


21 


89 


176 


167 


142 


66 


14 


8 


5 






691 




I 


26 


93 


197 


188 


105 


43 


14 


7 


3 


I 




678 


S 26 


I 


16 


81 


157 


163 


88 


46 


16 










568 


"0 ^7 


I 


21 


73 


90 


85 


58 


14 


7 


I 








350 


" 28 


2 


9 


33 


53 


54 


28 


9 


6 




I 






195 


29 


I 


8 


15 


22 


29 


7 


4 












86 


30 




2 


10 


10 


10 


7 














39 


31 






7 


3 


















10 


32 






I 




















I 


Totals 


10 


126 


491 


934 


909 


620 


274 


94 


29 


II 


I 


I 


3.500 



9. DENSITY OF COB AND WEIGHT OF GRAIN PER SECTION. 

It will be noted from Table 13 that the coefficient of variability of 
the density of cob is quite high, 17.51 + .1460. The extremes of 



212 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



density of cob range from 0.12 to 0.54 (factor). The examination 
of a large number of ears will show a wide variation in the density 
as well as the size (circumference) of cobs. What relation density 
of cob holds to the yield of grain is shown in Table 9. The coefficient 

Table 9. — Correlation between weight of shelled grain per section in grams, subject; and 
density of cob, relative, (r = — .0728 + .0113.) 



Density of cob (factor). 







00 












00 





0< 







00 












00 





N 








M 


M 




(N 


<N 


M 




cs 




ro 


ro 


ro 


ro 


■"^ 










10 


\n 




Totals 








1 


1 


1 


1 


1 


1 


1 


1 


1 


1 


1 




1 


1 


1 


1 


1 


1 








^• 


M 


M 




cs 


cs 




m 


ro 


ro 




ro 


% 








"t 


»o 


tr. 




I05-IIO 










I 










I 
























2 


IIO-II5 








I 




































I 


II5-I20 














































120-125 












2 




2 




























4 


125-130 














I 


I 




3 
























5 


130-135 












I 


I 




I 


I 






I 






I 












6 


135-140 






I 




I 


I 




2 


3 


I 


I 


5 




I 
















16 


140-145 












I 


I 


I 


4 


2 


3 






I 
















13 


145-150 












3 




2 


I 


8 


3 


4 




4 


5 














30 


150-155 








3 


3 


I 


I 


3 


6 


5 


7 


4 


5 


2 




I 


2 


I 






I 


45 


155-160 












3 


4 


9 


8 


12 


II 


II 


10 


2 


2 


4 












76 


160-165 










I 


4 


5 


8 


II 


II 


14 


9 


9 


I 


4 


5 


5 




2 


I 




90 


165-170 










2 


4 


10 


12 


17 


21 


21 


10 


19 


13 


4 


7 


4 


3 


I 






148 


170-175 










I 


3 


7 


22 


17 


19 


26 


18 


21 


18 


II 


8 


3 


I 


I 




I 


177 


175-180 










6 


5 


13 


24 


27 


26 


26 


30 


14 


24 


15 


9 


3 


5 




2 




230 


180-185 










4 


8 


9 


26 


24 


36 


24 


38 


26 


17 


12 


II 


5 






I 




241 


185-190 










I 


4 


23 


18 


27 


31 


34 


36 


33 


18 


12 


13 


9 


3 


2 






264 


190-195 








I 


3 


7 


14 


31 


34 


36 


38 


38 


20 


41 


13 


II 


I 


2 


2 






292 


195-200 










4 


10 


20 


32 


40 


38 


48 


32 


34 


21 


13 


10 


3 


5 








310 


200-205 






2 




6 


7 


21 


20 


46 


44 


53 


41 


38 


27 


10 


10 


5 


4 




3 




337 


205-210 


I 






I 


I 


6 


19 


24 


27 


34 


27 


37 


24 


29 


14 


8 


I 


3 




I 


I 


258 


210-215 








2 


3 


7 


13 


19 


36 


34 


26 


31 


22 


15 


6 


6 


3 


3 








226 


215-220 










5 


10 


17 


16 


24 


19 


34 


29 


22 


10 


7 


8 


I 


3 


2 




I 


208 


220-225 








I 


6 


8 


18 


26 


21 


19 


14 


9 


12 


4 


4 


3 


2 






I 




148 


225-230 








3 


I 


4 


8 


13 


14 


32 


25 


15 


12 


8 


5 


4 


3 










147 


230-235 










2 




4 


13 


7 


15 


9 


6 


7 


4 


I 


I 












69 


235-240 










2 




7 


7 


4 


8 


5 


5 


5 


3 


2 


I 












49 


240-245 












5 


2 


5 


8 


6 


4 


5 


I 


I 


I 


I 












39 


245-250 












3 


3 


5 


8 


I 


3 


2 


I 


I 
















27 


250-255 












I 


3 


I 


3 


I 


I 


I 




















II 


255-260 












I 


I 


I 


2 


5 


4 


2 


2 


I 




I 












20 


260-265 














2 






2 


I 






















S 


265-270 














I 






























I 


270-275 






























I 




I 










2 


275-280 




















2 
























2 


280-285 






















I 






















I 


Totals 


I 





3 


12 


53 


109 228 343 420 473 463 418 338 


266 


143 


123 


51 


33 


10 


9 


4 


3.500 



of correlation is low and negative, — .0728 + .0113. This indicates 
a very slight tendency for high-yielding ears to be associated with 
cobs of low density. 



GRANTHAM I EAR CHARACTERS IN CORN. 



213 



10. DENSITY OF COB AND WEIGHT OF KERNEL. 

An inspection of Table 10 shows a moderate negative correlation 
between the density of cob and the weight of kernel. The coefficient 
of correlation factor is — .1759 + .01 11. The relation between the 
two characters is to the effect that lighter kernels accompany the cobs 
of the greater density. It has been found in carrying on this work 



Table 10. — Correlation between weight of individual kernel {number in 10 grams) sub- 
ject; and density of cob, relative. (r = — .1759 + .01105.) 

Density of cob (factor). 





\0 00 







00 


CM 







00 









•0 


00 





CM 














W CO 
























10 


1 OLdlb 




1 1 1 

CM '^O 


CO 


1 




till 

M 00 







1 


1 


00 


1 




1 






00 





CM 










M oq M CM 


ro 


ro 


ro 


ro 


ro 




rf 








U-) 






18 








J 


























4 










2 


A 


2 






















12 


20 








3 I 


9 


8 


9 




4 


3 


2 


I 






I 




43 


2 I 








9 "2 Z A 
J 4 


1 




c 




10 


3 




3 






^ 




00 


22 








8 817 


18 


23 


23 


21 


19 





7 


2 


4 




I 


I 


lOI 


23 




I 




7 9 27 




29 


33 


34 


17 


7 


10 


4 


3 


I 


I 




229 


24 




I 


4 


6 39 


31 


41 


29 


30 


27 


10 


10 








I 


2 




296 


25 


I 




3 


9 14 16 28 


33 


62 


44 


33 


24 


22 


21 


7 


5 


2 




I 


325 


26 






9 


5 J-o 43 40 


45 


41 


UU 


41 


24 


17 


1 1 


7 


2 




I 


I 


371 


27 




I 


3 


II 32 25 45 


42 


52 


42 


30 


31 


19 


14 


4 


3 




2 




356 


zo 




2 


3 


7 25 30 33 


34 


42 


43 


22 


22 


9 


8 


6 


I 


I 






288 


29 


I 




A 
U 


11 zj 27 J7 


51 




23 


20 


20 


9 


Q 



I 




I 






277 




I 


3 


2 


9 18 32 19 


39 


30 


2 1 


2 1 


15 


6 


13 


2 


I 


2 






234 


31 




I 


I 


7 7 22 37 


30 


20 


22 


15 


12 


6 


4 


2 


I 


I 






188 


32 






3 


7 14 22 28 


28 


24 


20 


II 


4 


7 


6 


I 


4 








179 


33 






4 


5 12 17 18 


18 


13 


7 


II 


4 


5 


2 


2 










118 


34 


J 




c 



4 5 II 7 


13 


13 


15 


6 


5 


5 














95 


35 






2 


3 7 14 4 


12 


3 


4 


3 


3 


2 


2 


I 










59 


36 






I 


4 811 


10 


5 


7 


I 


5 




2 












54 


37 




I 


I 


4255 


6 


3 


2 


4 


I 


I 






2 








37 


38 








2661 


4 


6 


2 




2 
















29 


39 






2 


1233 


4 


2 


I 


3 


6 
















27 


40 








3 I I 




2 


2 








2 












II 


41 








2 I I 4 


2 


3 


I 


I 


I 
















16 


42 








I 22 


I 


I 


I 


2 


















10 


43 




I 


I 


I I 


I 


I 






















6 


44 








I 






I 




















2 


45 








I 


























I 


46 








I 


























I 


47 




































48 




































49 










I 
























I 


50 




































51 




































52 










2 
























2 


Totals 


103 


12 


53 


109 228 343 420 473 463 418 338 


266 


143 


123 


51 


33 


10 


9 


4 


3.500 



that the cobs of the greatest density are not the largest in circum- 
ference. The cobs of smallest circumference have a tendency to carry 
a smaller number of rows of kernels and these in turn are generally 
more shallow than on cobs of larger circumference. 



214 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



II. DENSITY OF COB AND THE DEPTH OF KERNEL. 

From the conclusion reached in the discussion of Table lo we 
should expect some correlation between the density of cob and depth 
of kernel. Large volume in cobs is not accompanied by the greatest 
density. Cobs of large circumference do not carry the largest kernels 
by weight. The coefficient of correlation between the density of cob 
and the depth of kernel is .1039 + .0113. The correlation is positive 
in a rather low degree, and indicates that depth of kernel is associated 
to some extent with density of cob. 

Table ii. — Correlation between density of cob, subject; and depth of kernel in 
centimeters, relative, (r — .1039 i -0113.) 



Depth of kernels in centimeters. 





00 

r 


r 
00 


q 

T 

o\ 


I.O-I.I 

1 


(N 

1 

M 


00 

IN 


1 


'it 







00 

1 


Totals 


.12-14 












I 












I 


.14-16 


























.16-18 






I 




I 


I 












3 


.18-20 




2 


2 


I 


2 


2 


2 


I 








12 


.20-22 


I 


I 


2 


8 


14 


13 


9 


4 




I 




53 


.22-24 




3 


3 


12 


33 


24 


21 


10 


3 






109 


.24-26 




3 


5 


29 


52 


58 


51 


23 


5 


I 


I 


228 


.26-28 






8 


38 


93 


88 


78 


34 


3 






343 


.28-30 






12 


50 


117 


114 


90 


28 


7 




I 


420 


.30-32 






13 


51 


121 


122 


90 


55 


18 


2 




473 


.32-34 






8 


42 


119 


125 


104 


51 


II 


2 




463 


•34-36 






7 


41 


97 


135 


78 


48 


9 


2 




418 


.36-38 






I 


37 


85 


83 


89 


38 


5 






338 


.38-40 






2 


20 


56 


87 


63 


30 


6 


2 




266 


.40-42 




I 


I 


12 


35 


51 


30 


8 


4 


I 




143 


.42-44 








12 


25 


36 


22 


21 


5 


2 




123 


.44-46 








3 


10 


19 


10 


8 


I 






SI 


.46-48 








I 


7 


II 


9 


4 


I 






33 


.48-50 










2 


4 


2 


I 


I 






10 


.50-52 










2 


2 


4 


I 








9 


.52-54 








I 






3 










4 




I 


15 


65 


358 


871 


976 


755 


365 


79 


13 


2 


3.500 



12. DENSITY OF COB AND THICKNESS OF KERNEL. 

It was found in Table 4 that the thickness of kernel was not 
directly correlated with the circumference of cob. In Table 8 it was 
shown that the weight of cob has no direct correlation with the 
thickness of kernel. Instead, the thicker kernels were found on cobs 
of moderate weight. The correlation between the density of cob and 
thickness of kernel is expressed by the factor — .0514 + .0114. It 
indicates a very slight tendency for thick kernels to be found on cobs 
of moderate density. 



GKAN'niAM: i"..\K ( ' 1 1 A l^\( "n:KS in corn. 



215 



Table 12. — Correlation hclwcrii density oj cob, subject ; and thickness of kernel 



(number in 10 centimeters) , relative. 


{r = 




0514 ± .0114.) 










Nuinbor 


of 


cci lie 


Is ill 


1 


(■(Mit iiiu'tcrs. 






18 


19 20 


21 


22 


23 


24 


25 


26 


27 


28 


29 30 31 32 


Totals 


.12-. 14 
















I 










.14-. 16 


























.16-. 18 


I 








I 


I 














.18-. 20 








2 




5 


3 


2 








1 2 


.20-. 22 




I 


3 


3 


9 


12 


5 


7 


9 


I 


I 2 


00 


.22-. 24 




I 2 


3 


6 


10 


16 


28 


22 


1 


Q 


2 I 


1 09 


.24-. 26 




3 


9 


19 


25 


50 


40 


30 


27 


15 


A -7 T 
031 


228 


.26-. 28 




2 2 





17 


43 


61 


69 


65 


30 


24 


13 2 I 


1A1 
J^O 


.28-.30 




4 


1 


32 


60 


75 


82 


66 


42 


30 


II 7 I 


420 


•30-. 32 




I 2 


15 


42 


58 


85 


84 


91 


30 


29 


15 9 3 I 


4/0 


•32-.34 


I 


I 5 


15 


30 


62 


lOI 


89 


63 


56 


28 


8 3 I 




•34--36 


I 


5 


10 


28 


65 


85 


80 


68 


40 


18 


10 7 I 


418 


•36-38 




6 


II 


32 


51 


72 


58 


40 


38 


17 


9 3 I 




•38-.40 




I 2 


5 


28 


34 


55 


58 


44 


25 


8 


5 I 


266 


.40-.42 




I 


3 


13 


22 


27 


30 


32 


I I 


3 


I 


143 


.42-.44 




I 


2 


8 


1 1 


20 


37 


21 


9 


9 


4 I 


123 


.44-.46 






3 


4 


9 


16 


7 


7 


3 


I 


I 


51 


.46-. 48 






3 


I 


. 8 


6 


3 


5 


3 


3 


I 


33 


.48-.50 










3 


I 


4 




I 


I 




10 


• 50-.52 




I 






3 


2 


I 


2 








9 


•52--54 




I 








I 




2 








4 



Totals! 3 7 35 98 265 474 691 678 568 350 195 86 39 10 



3.500 



Summary. 

From the data which have been presented on the correlation of 
characters of cob to other ear characters of corn the following con- 
clusions may be drawn : 

1. The yield of grain per ear is strongly correlated with the cir- 
cumference of cob. 

2. There is practically no correlation between weight of individual 
kernel and the circumference of cob. 

3. The depth of kernel is correlated to a moderate degree with 
cobs of small circumference. 

4. The thickness of kernel is slightly correlated with cobs of 
small circumference. 

5. The yield of grain per ear is correlated to a considerable degree 
with the weight of cob. 

6. The weight of kernel is moderately correlated with cobs of 
low weight. 

7. There is a very low correlation between depth of kernel and 
weight of cob. The heaviest cobs do not carry the deepest kernels. 

8. A fair degree of correlation exists between the thickness of 
kernel and cobs of low weight. 

9. The yield of grain per ear has a very slight correlation with low 
density of cob. 



2l6 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

10. There is a moderate degree of correlation between weight of 
kernel and cobs of low density. 

11. The depth of kernel is slightly correlated with density of cob. 



Table 13. — Summary table of variation constants in ears of corn and of corre- 
lation coefficients. 



The ear as the unit. 


Ex- 
tremes. 


Mean. 


Standard deviation. 


Coefficient of 
variation. 


Circumference of cob, cm 

Depth of kernel, cm 

Thickness of kernel (No. in 10 

cm.) 

Weight of kernel (No. in 10 

gm.) 


7-15 
15-75 
.12-.54 
105-285 
.7-1.9 

18-32 

18-52 


10.571 ±.0111 
36.500db.0830 
.328 ±.0006 
196.321 ±.2639 
I.248±.00I5 

24.827rfc.O229 

27.805rt.05i3 


0.9773 ±.0079 
.7289±.0590 
.0574rt.0004 

23.1466 rt. 1875 

.i345±-ooii 

2.0I37rt.0l63 

4.4994 ±-0364 


9.24:t.0748 

i9.97±-i68o 
17.51i.1460 
11.79i.0968 
10.78rfc.0883 

8.11 rfc.0657 

16.18dr.1344 


Characters. 


Coefficient of correlation. 


Circumference of cob and weight of grain per section 

Circumference of cob and weight of kernel 

Circumference of cob and depth of kernel 

Circumference of cob and thickness of kernel 

Weight of cob and weight of grain per section 

Weight of cob and weight of kernel 

Weight of cob and depth of kernel 

Weight of cob and thickness of kernel 

Density of cob and weight of grain per section 

Density of cob and depth of kernel 

Density of cob and thickness of kernel 


.4Il8rfc.0095 

— .Ol85rt.OII4 

— .I789rfc.0110 
-.I053rfc.OII3 

.3064dr.0i03 

— .l837rfc.0II0 

-.0747rfc.oii3 

— .I500rt.0III 

— .0728rfc.0II3 

-.i959±.oiii 
.i039±.oii3 

-.05I3rfc.0II4 



12. The correlation between thickness of kernel and density of 
cob is very low and negative. 

13. The coefficient of variability is much higher for weight of cob, 
density of cob, and weight of kernel than for the other characters. 

Literature Cited. 

Brigham, a. a. 

1896. Der Mais. Inaug. Diss. Gottingen. 
Craig, C. E. 

1908. Some correlations between physical characters of the maize plant, 
and between the physical characters and nitrogen content. Thesis 
submitted to the Faculty of Cornell University for the Degree of 
Master of Science in Agriculture. 
Davenport, E. 

1907. Principles of breeding, p. 460-467. Boston, Ginn & Co. 
EwiNG, E. C. 

1910. Correlation of characters in corn. N. Y. Cornell Agr. Expt. Sta. 
Bui. No. 287. 



Journal of the American Society of Agronomy. 



Plate 5. 




Fig. 3. View of plat planted with quartered 3-oiince potato tubers in 1916. 



Journal of the American Society of Agronomy. 



Plate 6. 




Fig. 3. View of plat planted with halved 4-ounce potato tubers in 1916. 



AICIUCR: WHOLK vs. cut POTATOKS for I'LANTING. 2 1/ 

Fruwirth, C. 

1904. Die Ziichtung dor landwirtscliaf tliclicn Kulturpflanzen, 2: 8-13. 
Berlin, Paul Parcy. 

Thiele, p. 

1899. Dor Maisbau. Stuttgart, Eugen Ulmer. 
Vries, H. de. 

1901. Die Mutationstheorie. Leipsic, Vert & Co. 
Delaware Agricultural Experiment Station, 
Newark, Delaware. 



WHOLE VS. CUT POTATO TUBERS FOR PLANTING ON 
IRRIGATED LAND.— i.^ 

L. C. AlCHER. 

In the United States practices in potato production are very largely 
sectional and often differ greatly. Probably the greatest difference is 
noted in the size and nature of the tuber piece planted. In some 
sections, the whole potato is planted ; in others, only cut tubers. 
Some growers plant a large set, others plant a small one, and some 
plant culls. Only two of the great potato-producing sections, so far 
as the writer is able to learn, grow the same variety as a major crop. 
Climate and length of growing season are large determining factors 
in potato varietal selection and may also be contributing causes of 
some of the above mentioned methods of production. 

There is considerable printed information available on potato 
production. Information tending to establish rather definitely the 
size of tuber piece to plant for the most economical production of 
the Irish potato, however, is limited and sectional. Practically all the 
information which is available has to do with potato production under 
humid or subhumid conditions. The experiments here reported deal 
entirely with the potato under irrigation. 

These experiments were begun on the Aberdeen Branch Experi- 
ment Station at Aberdeen, Idaho, in 1913. In 1912 suitable seed 
stock was grown for the work. As the station work became better 
organized, more time and money were spent in obtaining data which 

1 Contribution from the Idaho Agricultural Experiment Station. Received 
for publication February 9, 1917. This is the first of two papers on the use 
of whole or cut potato tubers of various weights for planting on irrigated land, 
with reference to the total yield and the yield of marketable tubers. The work 
here described was performed on the Aberdeen substation. The second paper, 
which appears elsewhere in this issue, describes similar experiments conducted 
at the Gooding substation by J. S. Welch. 



2l8 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

at first were not taken. It is realized that still further information 
would be highly desirable. 

Outline of the Experiment, 

The Idaho Rural potato was used in this experiment. This variety 
is now supposed to be of the same stock as the Charles Downing. It 
it a flattened, oval-oblong, very smooth, medium-sized tuber with 
shallow eyes and white, creamy skin. The sprout is white, with tips 
slightly colored. 

The set, or seed piece, as the terms are here used, is the portion of 
the tuber used in planting. The experiment was planned to include 
maximum, optimum, and minimum sized seed pieces. Tubers weigh- 
ing approximately 8 ounces, 4 ounces, and 3 ounces each were used. 
A portion of each lot was planted whole, others were halved, and 
still others were quartered. The halved pieces were obtained by cut- 
ting the tuber lengthwise, care being taken to divide it into equal 
parts. The quartered seed was obtained by cutting the tuber length- 
wise and then crosswise, making the quarters as uniform in size as 
possible. No attempt was made to have a definite number of eyes 
on each piece. Each whole tuber was first examined for external 
indications of disease. A thin slice then was cut ofif the stem end of 
the tuber as a further precaution against planting diseased stock. 
All tubers showing brown discoloration were discarded. 

In 1913 and 1914 the seed potatoes were given the formaldehyde 
treatment (i pint of formaldehyde to 25 gallons of water). The 
tubers were soaked in this solution for two hours. In 191 5 and 1916 
the bichloride treatment was given (4 ounces of bichloride of mer- 
cury to 30 gallons of water). The tubers were soaked i^ hours. 
Those which were to be planted whole were treated after the stem 
ends had been examined for discoloration. Tubers which were to be 
cut were treated after cutting, to avoid reinfection of disease if any 
was met in cutting. 

Care was taken to get tubers of the proper size for each experi- 
ment. A variation of a half ounce was permitted in the selection of 
the 8-ounce lot, while a variation of but a quarter ounce was per- 
mitted in the selection of the 3-ounce and 4-ounce tubers. The aver- 
age weights in each lot were rigidly maintained. 

The soil on the Aberdeen substation is a sandy clay loam. The 
191 3 crop was planted on land which was cleared of sagebrush two 
years previous and had produced two successive crops of wheat. 
The 1914 crop was planted on land which had produced three sue- 



aiciikk: wnoLK vs. err potatoes for plantinc. 219 

cessive crops of wheat and which was niaiuired in the fall before the 
potatoes were planted. The 1915 crop was planted on land which 
})roduced a crop of held peas harvested for seed and which had pre- 
viously been in wheat for three successive years. This land was 
manured the fall before the peas were planted. The 191 6 crop was 
planted on land that had been in alfalfa three years and previous to 
that in wdieat two years. 

Furrows were opened with a shovel plow and the sets dropped by 
hand and covered with a " crowder." The rows were 33 inches apart, 
6 rows composing the unit of a twentieth acre. The sets were planted 
16 inches apart in the row and 4 inches deep. A brass chain with 
strings tied at 16-inch intervals served as a guide to the planters, 
thus insuring accurate placement of each tuber set. This method of 
planting proved very satisfactory. 

Three or four irrigations and cultivations were given during the 
growing season, the number of irrigations given depending upon the 
seasonal precipitation and other climatic factors. Cultivation always 
followed irrigation except when the vines were large enough to 
shade the ground. 

In 191 5 and 1916 counts were taken on the percentage of stand and 
number of stalks per hill. 

The yields of marketable and cull tubers were taken each year 
during the 4-year period. These data were obtained by hand sorting 
in 1913 and sorting over a specially prepared 2-inch screen in 1914, 
1915, and 1916. In connection with this part of the experiment, the 
number of tubers in a bushel of the marketable and a bushel of the 
cull tubers was determined in 191 5 and 191 6, in order to arrive at a 
definite conclusion relative to the sizes of tubers produced by the 
various sizes of tuber sets and their average weight. 

The crop was harvested with a standard type of power potato 
digger. Each plat was stored in the cellar until all were dug. The 
various lots were then sorted and calculations made. 

Results. 

The sprouting and emergence period varied almost directly with 
the size and kind of set planted. The plants from the 8-ounce, 
4-ounce, and 3-ounce whole tubers appeared first and in the order 
here mentioned. The next plants to appear came from the 8-ounce 
halved tubers. The plants from the 4-ounce and 3-ounce halved and 
8-ounce quartered tuber sets appeared in the next 24 to 48 hours. 
The plants from the 4-ounce and 3-ounce quartered tubers were the 



2 20 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

last to appear, the latter appearing from two to four days later than 
the plants from the 8-ounce whole tubers. 

The sprouting and emergence period varied considerably from year 
to year, depending upon climatic conditions and the condition of the 
soil. A moist, warm soil was conducive to quick growth. A dry, 
warm soil retarded the growth of the smaller cut-tuber pieces but 
only slightly affected the emergence of the plants from the whole 
tubers. The early growth of the plants from the large whole tubers 
was least aft'ected by the soil moisture content and the physical con- 
dition of the soil. Cold soil and cold weather retarded the growth 
of the plants from the whole and the cut tubers, but the latter seemed 
most affected. 

Table i shows that the stand from the whole and halved tubers 
was excellent. There was, however, a severe loss in stand from the 
quartered tuber sets. The loss in stand from the 8-ounce and 4-ounce 
quartered tubers was 11 percent, while the loss from the 3-ounce 
quartered tubers was almost 18 percent. This loss is attributed to 
the inability of the plants from the small tuber sets to overcome adverse 
climatic and soil conditions, and to a possible lack of eyes on many 
of the small sets. The Idaho Rural potato has from 8 to 10 eyes, 
rarely 11, and 6 or more of these eyes are on the bud-end half of 
the potato. When this tuber is quartered without regard to the 
placement of the eyes, the stem-end quarters may have one or more 
eyes or sometimes none at all. 

The average number of stalks per hill, as determined by actual 
count, varied directly with the size of the tuber set, in each respective 
lot of whole, halved, and quartered tubers. The 4-ounce whole tubers 
produced a greater number of stalks per hill than the 8-ounce halved 
tubers, which are really 4-ounce sets. This is due to the greater 
number of eyes on the whole tuber. The 4-ounce halved set produced 
more stalks per hill than the 8-ounce quartered tubers, for the same 
reason, both being 2-ounce sets. The 4-ounce quartered set produced 
the same number of stalks as the 3-ounce quartered tuber. This is 
due to the sets being very nearly the same size and having about the 
same number of eyes. Table i shows that the whole tubers produced 
almost twice as many stalks per hill as the halved sets and that the 
halved sets produced almost twice as many stalks per hill as the 
quartered sets. 

Practically every eye on the 8-ounce whole, halved, and quartered 
tuber sets produced stalks, as will be noted by comparing the number 
of stalks per hill and the number of eyes per tuber on the Idaho 



aiciier: wiior.K vs. cut i»()Tat()i:s for plantinc; 



22 I 



Rural potato. The earlier si)routin^ bud-eye on the larger sets does 
not ^row to the exclusion of j^rowth of the other eyes of the tuber. 
The Idaho Rural potato is not ^ivcn to producing a master sprout. 
If this potato, by breeding or selection, could be made to grow a 
master s])rout the undesirable features resulting from planting whole 
tubers would be eliminated. The size of the set determines very 
largely how many eyes will produce stalks, for the larger the set 
the more plant food there is available and the greater the stimulus 
to each eye to produce a plant. The 4-ounce and 3-ounce whole 
tubers have just as many eyes as the 8-ounce tubers, but the eyes on 
the small tubers are not as well developed. Table i shows that the 
8-ounce whole tubers produced 8.67 stalks per hill, whereas the 
4-ounce and 3-ounce whole tubers produced but 5.41 and 4.82 stalks 
per hill, respectively. These stalks produce tubers and it is this 
fact which has such a large bearing upon the percentage of market- 
able tubers from the whole, halved, and quartered seed. These per- 
centages are given in Table i. 

Table i. — Average stand, number of stalks per hill, total yield, yield of mar- 
ketable tubers, and size of marketable and cull tubers recorded in an experi- 
ment to determine the effect of planting zvhole or cut potatoes of various sizes 
on irrigated land at the Aberdeen substation, Aberdeen, Idaho, iqij to 1916, 
inclusive. 



Description of 
tuber set planted. 

8-oz. whole . . . 
8-oz. halved . . 
8-oz. quartered 
4-0Z. whole . . . 
4-0Z. halved . . 
4-0Z. quartered 
3-0Z. whole . . . 
3-0Z. halved . . 
3-0Z. quartered 


Stand. a 


Stalks 
per 
hill.« 


Yield per acre. 


Percent- 
age of 
market- 
able 
tubers. 


Number of tu- 
bers per bushel." 


Weight per tuber." 


Total. 


Market- 
able. 


Market- 
able. 


Culls. 


Market- 
able. 


Culls. 


Percent. 
99.91 
99-97 
89.28 
99.99 
99.99 

89-31 
100.00 
98.87 
82.19 


8.67 
4.71 
2.63 
5-41 
2.98 
1. 71 
4.82 
2.64 
1.72 


Bushels. 
392-9 
333-5 
314.0 
368.7 
332.9 
322.7 
361-7 
355-5 
262.7 


Bushels. 
200.6 
210.5 
218.2 
171. 
220.1 
250.9 

201. 1 
253-8 
201.5 


52.6 
65.2 
69.1 
46.3 
66.1 
77-4 
54-2 
68.8 
78.0 


209 
196 
152 
179 
171 
152 
196 
162 
170 


504 
439 
410 

463 
418 

414 
449 
417 
412 


Ounces. 

4.6 
4.9 

5-7 
5-3 
5-7 
6.2 
4.9 
5-9 
5-6 


Ounces. 
. 1.8 

2.1 

2.3 

2.0 

2.2 

2.3 
2.1 

2.3 
2.2 


Summary. 


Whole 

Halved 

Quartered .... 


99-97 
99.61 
86.92 


6.30 

3-44 
2.02 


374-4 
340.6 
299.8 


191. 

228.1 
223.5 


51.0 
66.7 
71-5 


194 
176 
158 


472 
424 
412 


4- 9 

5- 5 
5-8 


1.9 
2.2 
2.2 



"Average for two years only (1915 and 1916). 



Whole tubers invariably produced a larger and more plentiful 
growth of top than the cut pieces. This growth was proportional to 



222 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

the size of the whole, halved, and quartered sets, respectively. 
Plates 5 and 6 show clearly the variations in growth. The large and 
vigorous growth from the large sets is due to the better start given 
the plants by the greater supply of food available ; to the greater 
proportion of eyes producing stalks ; and to the fact that the greater 
supply of natural moisture in the larger sets is not exhausted so 
rapidly. Unless the soil is in good condition, the small set dries out 
before the plant can get sufficient moisture from the soil to sustain 
growth. Some of the loss in stand from the quartered small tuber 
sets was due to unfavorable soil conditions. The top growth from 
the 3-ounce quartered sets was the least vigorous and the number 
of stalks per hill was the lowest of all the lots. 

The whole tuber sets in each lot invariably produced the largest 
total yield of potatoes per acre. The average difference in yield 
between the whole and the halved tuber sets was 33.8 bushels per 
acre, while the average decrease in yield from halved to quartered 
sets was 40.8 bushels per acre. The whole-tuber plats outyielded the 
quartered-set plats on the average by 74.6 bushels per acre. The 
total yield from the 8-ounce whole tubers averages 28 bushels per 
acre more than the yield from the 4-ounce and 3-ounce whole tubers. 
On the other hand, the whole tuber sets have invariably yielded the 
smallest percentage of marketable potatoes per acre. 

The total yield per acre, the yield per acre of marketable tubers, 
and the percentage of marketable tubers recorded in Table i are 
4-year averages. The 8-ounce whole tubers produced 52.6 percent of 
marketable tubers, whereas 78 percent of the crop from the 3-ounce 
quartered sets was marketable. The 3-ounce quartered sets have 
always produced the smallest yield in bushels per acre and the highest 
percentage of marketable tubers. The decrease in yield is due in 
great measure to the 18 percent loss in stand in the 3-ounce quartered 
plats. With this great handicap, the 3-ounce quartered sets main- 
tained a small increase in yield per acre of marketable tubers over 
the 8-ounce, 4-ounce and 3-ounce whole sets. The 8-ounce and 
4-ounce halved and quartered sets and the 3-ounce halved sets all 
yielded a much greater number of bushels of marketable tubers per 
acre than any of the lots of whole seed or the 3-ounce quartered sets. 
The percentage of marketable tubers increased as the size of the set 
decreased. This is due in part to the smaller number of stalks per 
hill, to the smaller numbers of tubers per hill, and to the greater 
growth per tuber in the hills from the smallest sets. Whole sets 
produced more tubers per hill than cut sets and the average size of 



AiciiKu: vviioLi-: vs. cut potatoes for planting. 



223 



the tuber was much smaller. I he larger the set the greater the 
number of tubers produced and the smaller the average weight per 
tuber. This is due to the greater number of stalks per hill from the 
large sets, the greater competition for moisture and food, and the 
greater number of tubers produced per hill. 

The data recorded in the experiment are summarized in Table i. 

Summary 

1. Whole tuber sets sprouted and the plants came up more quickly 
than those from cut tubers. 

2. Whole tuber sets produced a larger and more plentiful top 
growth than cut tuber sets. 

3. The vigor and size of the plant increased as the size of the set 
increased. 

4. The number of stalks per hill increased directly as the size of 
the set increased. 

5. The loss in stand from planting whole and halved tubers aver- 
aged less than i percent, while the loss in stand from planting quar- 
tered tubers averaged almost 13 percent. 

6. The earlier sprouting bud eye of the Idaho Rural potato does 
not grow to the exclusion of the other eyes of the tuber. 

7. The total yield from whole tubers was 14.4 percent more than 
from cut tubers. 

8. Cut tubers yielded 18 percent more marketable potatoes per 
acre than whole tubers. 

9. The percentage of marketable tubers increased as the size of 
the set decreased. 

10. The larger the set the greater was the number of tubers pro- 
duced and the smaller the average weight per tuber. 

Aberdeen Substation, 
Aberdeen, Idaho. 



224 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



WHOLE VS. CUT POTATO TUBERS FOR PLANTING ON 
IRRIGATED LAND.— 2.^ 

John S. Welch. 

The production of potatoes on the irrigated lands of southern 
Idaho is an extensive and profitable industry. Growers generally arc 
agreed upon certain methods of procedure in the production of 
potatoes under irrigation, but there is a widespread difference of 
opinion among them relative to the use of whole and cut tubers for 
planting. Prominent growers unqualifiedly recommend the planting 
of whole tubers and argue that such a procedure results in better 
stands, earlier development, greater yields, and greater profits. 
Other growers equally as prominent reason that the old practice of 
cutting the tubers into a number of pieces is productive of best 
results under the conditions which prevail on most irrigated farms. 

Review of the Literature. 

In reviewing the literature of the subject it seems best to mention 
only conclusions based directly on experimental data. 

S. Johnson^ concluded after four years of experimental work that 
whole tubers produce a greater total yield and also a greater yield of 
unmarketable potatoes than cut tubers. 

D. D. Johnson^ found that whole tubers produced a greater num- 
ber of vigorous stalks per hill than halved or quartered tubers, but 
that the increased number of stalks did not produce a relatively 
greater yield of potatoes. 

Taft* grew two varieties of potatoes from tubers planted whole and 
cut into halves, quarters, and eighths. He found halved tubers to be 
the most desirable when considered from the standpoint of net profit 
per acre. 

1 Contribution from the Idaho Agricultural Experiment Station. Received 
for publication February 2i, 1917. 

2 Johnson, Sam'l. Potatoes, roots, fertilizers and oats. Mich. Agr, Expt. 
Sta. Bui. 46. 1889. 

3 Johnson, D. D. Potato culture and fertilization. W. Va. Agr. Expt. Sta. 
Bui. 20. 1892. 

4 Taft, L. R. Potato tests. Mich. Agr. Expt. Sta. Bui. 85. 1892. 



WKLCII : WIIOLK VS. CUT l'()TAT()i:S KOU I'LANTINCi. 225 

I larwood and Holden'^ grew four varieties in an experiment to 
determine the relative value of whole and cut tubers for ])lanting. 
With three of the varieties, plantinj^ whole tubers produced the 
greater yield of marketable potatoes. In every case the whole tuljcrs 
produced the greater yield of unmarketable potatoes. 

I^lumb" found that planting halved tubers ])roduced a greater num- 
ber and a greater weight of marketable i)otatoes per hill than planting 
whole tubers ; that whole tubers produced nearly twice as many 
unmarketable potatoes per hill as the halved tubers ; that whole tubers 
produced a greater total weight of potatoes per hill ; and that the 
potatoes grown from halved tubers were of a greater average size. 

Experimental Data. 

In order to obtain data that would help to settle questions that 
arise in the irrigated sections of Idaho relative to the advisability of 
using whole or cut potatoes for planting, the following experimental 
work was conducted at the Gooding Substation during 1914, 191 5, 
and 1916. Because of the uniformity of the results obtained, it is 
thought necessary to present in this paper only the average results of 
the three years' work. 

Each season the test was conducted on eight plats of two rows 
each, with an average of 1 10 hills per row. The sizes of tuber pieces 
planted were as follows: 8-ounce* to lo-ounce tubers, whole, halved, 
and quartered ; 4-ounce to 6-ounce tubers, whole, halved, and quar- 
tered ; and 2-ounce to 3-ounce tubers, whole and halved. 

The soil on which the experiment was conducted is a uniform 
medium clay loam. Previous to the summer of 1909 it was covered 
with a rank growth of sagebrush. After being cleared and before 
being cropped to potatoes, it had produced small grains and legumes 
and had been given applications of barnyard manure. In preparation 
for potato growing the land was fall-plowed and left rough over 
winter. In the early spring it was worked down to conserve as much 
as possible of the winter precipitation until planting time. 

The Idaho Rural variety was used exclusively. Stuart' classifies 
this variety with the Green Mountain group, which he describes as 
follows : 

5 Harwood, P. M., and Holden P. G. Potatoes : Amounts of seed. Mich. 
Agr. Expt. Sta. Bui. 93. 1893. 

^ Plumb, C. S. Experiments in growing potatoes. Tenn. Agr. Expt. Sta. 
Bui., vol. 3, no. I. 1890. 

Stuart, William. Group classification and varietal descriptions of some 
American potatoes. U. S. Dept. Agr. Bui. 176. 1915. 



2 26 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Vines large, strong, healthy and well branched. Stems nearly upright in 
early stages of growth but gradually assuming a spreading habit toward the 
latter end of«the season. Flowers white, abundant, rarely producing seed balls 
except under very favorable soil and climatic conditions. Tubers broadly 
roundish-flattened to distinctly oblong-flattened; ends usually blunt, especially 
the seed end ; eyes medium in number, rather shallow with strong bud-eye 
cluster; skin dull creamy white, more or less netted; frequently with russet- 
colored splashes toward the seed end; sprouts rather short and stubby. 

The tubers used for planting were bright, well-kept stock and true 
to type. The selections for size were made by weighing. In halving, 
the tubers were cut lengthwise (from seed end to stem end) ; in quar- 
tering they were first cut lengthwise and then crosswise. Planting 
was done by hand immediately after cutting. The sets were planted 
in hills i8 to 20 inches apart in rows 3.5 feet apart. 

All plats were cultivated and irrigated exactly alike. The water 
was applied in a small stream running in a comparatively deep furrow 
between rows. On the average, three irrigations per season were 
given. All harvesting was done by hand. Potatoes which passed 
through a 2-inch mesh screen were classed as culls. The number of 
stalks and tubers per hill was estimated from data obtained by count- 
ing carefully the plants and tubers in 50 hills of each plat. All 
records in the tables are from actual weighings. The rates of 
planting, the stands obtained, and the number of days from planting 
to emergence are shown in Table i. 



Table i, — Rate of planting, stands, and number of days from planting to emer- 
gence recorded in an experiment to determine the relative merits of tubers 
of various sizes and of whole and cut tubers for planting. 



Size and portion of tuber planted. 


Rate of planting 
per acre. 


Stand. 


Time from planting 
to emergence. 




Pounds. 


Percent. 


Days. 


8 to 10 ounces, whole 


5.000 


93-28 


24 


8 to 10 ounces, halved 


2,500 


95-69 


24 


8 to 10 ounces, quartered 


1,250 


87-34 


27 


4 to 6 ounces, whole 


2,800 


96.06 


24 


4 to 6 ounces, halved 


1,400 


91.22 


27 


4 to 6 ounces, quartered 


700 


90.39 


28 


2 to 3 ounces, whole 


1,400 


93-95 


27 




700 


92.50 


28 



The cost of producing a crop of potatoes must be taken into account 
in ascertaining the relative efficiency of production methods. As the 
price of good seed potatoes at planting time is usually high, their cost 
becomes one of the most important factors in determining the cost of . 
production. The number of pounds of the various seed pieces 



WELCH: wjioij-: vs. cut I'orAi'oics for i'lantinc. 227 

required to plant an acre is, therefore, sliown in 'l<'il)le 1. Since 
Q,ooo hills per acre are usually i)lante(l on the irrigated lands of tliis 
section, that numher was used in making- the computations. It is 
clear that in computing- the net returns per acre, especially from 
planting- whole tuhers, the quantity planted per acre nuist be 
reckoned with. 

Except in the case of the potatoes weighing 8 to 10 ounces, the 
whole tubers produced a slightly better stand than the halved tubers ; 
in all cases the halved tubers produced a better stand than the quar- 
tered tubers. 

One of the arguments used by those who advocate the planting of 
whole tubers is that this procedure results in a shortening of the time 
intervening between planting and emergence and thus insures the 
earlier development of the crop. It is apparent from the data 
recorded in Table i that the time from planting to emergence depends 
entirely on the size of the seed piece, and not at all on whether it is 
whole-) halved, or quartered. 



Table 2. — Numher of stalks, numher of tuhers, and numher of marketable 
tuhers and culls per hill produced from tuhers of various sizes planted 
whole, halved, and quartered. 



ft 



Size and portion of tuber planted. 


Number of 
stalks per 
hill. 


Number of 
tubers per 
hill. 


Number of 
tubers per 
stalk. 


Number of 
marketable 
tubers per 
hill. 


Number of 
culls per 
hill. 


8 to 10 ounces, whole 


8.9 


22.7 


2.6 


7-3 


15.4 


8 to 10 ounces, halved 


5-6 


17.2 


3-1 


8.2 


9.0 




2.9 


12. 5 


4-3 


7.2 


5-3 


4 to 6 ounces, whole 


7-4 


20.8 


2.8 


8.1 


12.7 


4 to 6 ounces, halved 


4.0 


15.0 


3-7 


8.0 


7.0 




2.3 


12.0 


5-2 


7-4 


4.6 


2 to 3 ounces, whole 


5-2 


16.4 


3-2 


7.6 


8.8 


2 to 3 ounces, halved 


2.9 


12.4 


4-3 


6.4 


6.0 



Recommendations for the use of whole tubers for planting seem 
also to be based in part upon the belief that, when the tuber is planted 
whole, one large sprout will grow from the terminal bud or seed end 
and will develop at the expense of other sprouts, producing one large, 
vigorous stalk per hill. Data recorded in Table 2 do not support 
that belief. Nearly all the eyes grow. The whole tubers produced 
nearly twice as many stalks per hill as the halved tubers and the 
halved tubers nearly twice as many stalks as the quartered tubers. 
The number of tubers produced per hill decreased with the number 
of stalks per hill, although the decrease was not directly in propor- 
tion. The number of tubers per stalk increased as the number of 



228 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Stalks per hill decreased. The number of cull potatoes per hill was 
in almost direct proportion to the number of stalks per hill. 

As would be expected from the data shown in Table 2, the plant- 
ings which produced the fewest culls produced tubers of greatest 
average size. In every size of tubers planted (8- to lo-ounce, 
4- to 6-ounce, and 2- to 3-ounce), the whole tubers produced smaller 
potatoes than the halved and the halved tubers produced smaller 
potatoes than the quartered. 



Table 3. — Average weight of all tubers, and of, marketable tubers and culls 
separately, produced from planting tubers of various sizes planted 
whole, halved, and quartered. 



Size and portion of tuber 
planted 


Average size of tubers. 


Weight of tubers per hill. 


Percent- 
ape 
market- 
able. 


All. 


Market- 
able. 


Culls. 


All. 


Market- 
able. 


Culls. 




Ounces. 


Ounces. 


Ounces. 


Pounds. 


Pounds. 


Pounds. 




8 to 10 ounces, whole . . . 


2.6 


4-7 


1.4 


3-67 


2.28 


1-39 


62.12 


8 to 10 ounces, halved . . 


3-5 


5-7 


1.6 


3-76 


2.90 


.86 


77.12 


8 to 10 ounces, quartered 


4-3 


5-9 


2.2 


3.36 


2.65 


• 71 


78.87 


4 to 6 ounces, whole .... 


2.8 


4.6 


1.6 


3.56 


2.33 


1.23 


65-45 


4 to 6 ounces, halved . . . 


3-5 


5-0 


2.0 


3-33 


2.50 


.83 


75-05 


4 to 6 ounces, quartered. 


3.8 


4.8 


2.1 


2.85 


2.23 


.62 


78.25 


2 to 3 ounces, whole .... 


3-2 


4-7 


1.9 


3-25 


2.24 


1. 01 


68.09 


2 to 3 ounces, halved . . . 


4.0 


6.0 


1.9 


3-II 


2.41 


.70 


77-49 



Any practice in potato culture should be judged according to the 
yield of marketable potatoes obtained by following it. It will be 
noted that 8-ounce to lo-ounce whole tubers produced a smaller 
yield of marketable potatoes per hill than either the halved or the 
quartered tubers and that the halved tubers produced a greater yield 
of marketable potatoes per hill than the quartered ones. The whole 
tubers weighing 4 to 6 ounces produced a smaller yield of marketable 
potatoes than the halved, but a slightly larger yield than the quar- 
tered tubers. Even with tubers weighing 2 to 3 ounces, the halved 
outyielded the whole tubers in yield of marketable potatoes. 

From the standpoint of total yield, the greatest weight was grown 
from 8- to lo-ounce tubers halved, the next greatest weight from 
8- to lo-ounce whole tubers, and the next from 4- to 6-ounce whole 
tubers. The smallest total yield was produced from 4- to 6-ounce 
tubers quartered. 

The yields are shown in Table 4 in pounds per hill. On that basis 
the variations do not seem great. When it is remembered, however, 
that a difference of i pound per hill is equivalent to about 9,000 
pounds per acre, these seemingly small variations become very sig- 
nificant. 



WKLcii: W1IOLI-: vs. cut totatok^ for I'LANTiNc. 229 

The more important data in Tables 2 and 3 are summarized in 
Table 4 so that they may be more eonveniently studied. 



Table 4. — Sumviary of principal data on an experiment with planting whole 
and cut tubers of various sizes. 



Size and portion of tuber planted. 


Number of 
stalks per 
hill. 


Number of 
tubers per 
hill. 


Average 
weight 
of tubers. 


Total 
weight 
of tubers 
per hill. 


Total 

weight of 
marketable 
tubers 
per hill. 


Percentage 
of tubers 
market- 
able. 








Ounces. 


Pounds. 


Pounds. 




8 to 10 ounces, whole 


8.9 


22.7 


2.6 


3.67 


2.28 


62.12 


8 to 10 ounces, halved 


5-6 


17.2 


3-5 


3-76 


2.90 


77.12 


8 to 10 ounces, quartered . . . 


2.9 


12.5 


4-3 


3.36 


2.65 


78.87 


4 to 6 ounces, whole 


7-4 


20.8 


2.8 


3-56 


2.33 


65-45 


4 to 6 ounces, halved 


4.0 


15.0 


3-5 


3-33 


2.50 


75-05 


4 to 6 ounces, quartered .... 


2.3 


12.0 


3-8 


2.85 


2.23 


78.25 


2 to 3 ounces, whole 


5-2 


16.4 


3-2 


3-25 


2.24 


68.09 


2 to 3 ounces, halved 


2.9 


12.4 


4.0 


311 


2.41 


77-49 



It appears that a definite relation exists between the number of 
Stalks per hill, the number of tubers per hill, the average size of the 
tubers, and the percentage of marketable tubers ; that the number of 
tubers per hill decreases with the number of stalks per hill and that 
this decrease is accompanied by an increase in the average size of the 
tubers and in the percentage of marketable tubers. 

The best yields were obtained with a medium number of tubers 
per hill. It is conceivable that, if sufficient plant food were available, 
all of the tubers which set might be made to reach marketable size. 
In that case hills which set the greatest number of tubers would 
produce the greatest yields. If this condition could be realized one 
could assume that the procedure which induced the greatest set per 
hill would be the most profitable. A hill of eight stalks has little 
if any greater root zone than one of two stalks and, therefore, it 
has no more plant food to draw upon. To provide that hill with suffi- 
cient plant food to develop all of its tubers, a very liberal use of 
commercial fertilizer would be necessary. Under present conditions 
on the average irrigated farm in this section it is doubtful if that 
practice would prove practicable. It must not be inferred, however, 
that the soil upon which this experiment was conducted was in a low 
state of fertility, for potatoes of the same variety growing within a 
few feet of the plats used in 1916 yielded 450 bushels per acre. 

Most of the tubers produced on the experimental plats showed a 
small amount of rhizoctonia. Tubers from each plat were carefully 
examined to determine the relative amount of disease present. No 



230 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

difference whatever could be noted between the potatoes grown from 
whole and those from cut tubers. 

Summary. 

The experiments recorded here have been conducted for three 
years. The greatest possible care was taken to insure uniform soil 
conditions and to provide uniform irrigation and cultural methods. 
The results of the three years' work are in close agreement. Average 
results only are presented in this paper. 

In general, planting whole tubers produced a better stand than cut 
tubers, but the increase in stand was not at all commensurate with 
the greater weight of tubers used in obtaining it. 

The relative time from planting to emergence did not depend upon 
the use of whole or cut tubers. It was determined wholly by the 
size of the seed piece. 

Whole tubers showed no tendency whatever to develop only one 
sprout from the seed end. 

The number of tubers per hill decreased with the number of stalks 
per hill, but the number of tubers per plant increased as the number 
of stalks per hill decreased. The hills which had the greatest number 
of stalks invariably produced the greatest number of culls. 

In every size planted, whole tubers produced smaller potatoes than 
halved and halved produced smaller potatoes than quartered tubers. 

The greatest yield of marketable potatoes was produced from 8- 
to lo-ounce tubers halved. Very good yields, however, were pro- 
duced by the 8-to lo-ounce quartered and the 4- to 6-ounce halved 
tubers. In every weight of tubers selected the whole tubers produced 
smaller yields of marketable potatoes than the cut tubers. 

Under conditions which prevail on the average irrigated farm in 
southern Idaho, the planting of whole potatoes is not advisable. 

Gooding Substation, 
Gooding, Idaho. 



Journal of the American Society of Agronomy. Plate 7. 




bosiinakian: souAui:iii:Anr.i)Ni:ss in wiikat. 



THE COMPARATIVE EFFICIENCY OF INDEXES OF DENSITY, 
AND A NEW COEFFICIENT FOR MEASURING 
SQUAREHEADEDNESS IN WHEATJ 

S. Bosiinakian. 

« 

There are three types of compact wheats, the squarehead (Triti- 
cum capitatum, Schulz.), the club (T. compacHim, Host.) and a 
third form, the squarehead-ckib, which will be referred to in this 
paper as T. compact o-capitatum. In figure 14 these three forms are 
represented by heads 4, 5, and 6, respectively. These are all classi- 
fied generally as T. compactum and oftener called " club " in litera- 
ture, but as they differ appreciably in form as well as in genetic 
behavior it is necessary to make distinctions between them. The 
object of this paper is first to analyze the comparative efficiency of 
the indexes of density in use at present ; second, to suggest the use 
of a new coefficient to substitute for the present ways of measuring 
compactness, which do not bring out these differences ; and third, to 
describe an instrument for measuring squareheadedness. 

I. The Indexes of Density. 

The index of compactness, known as the density coefficient, may 
be determined in several ways. The oldest in use in practical breeding 
was found by the formula 

Di = (Formula la) 

where density according to formula i; 

L = length of spike measured from the base of the head to the tip 
of the terminal spikelet ; and 
= number of spikelets. 

The length of the head measured in this manner will vary accord- 
ing to the length of the terminal spikelet. When, for instance, 
density studies are to be made upon a population derived from a 
cross between a Polish and any of the club wheats, the use of this 
formula becomes anything but practicable, due to the length of the 
terminal spikelet of the Polish, which sometimes is as long as the 

1 Paper No. 61, Department of Plant Breeding, Cornell University, Ithaca, 
N. Y. Received for publication January 10, 1917. 



232 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



entire rachis of the club wheat. In taking data on vulgare wheats the 
use of formula i, as will be pointed out later, does not introduce a 
very serious error. 

To overcome this source of error another method came into use. 
By this method the index of density is found by dividing the length 
of the rachis (instead of the total length of the head) by the number 
of spikelets, thus : 

£>2 = — ' (Formula 2a) 

where = density found by formula 2a; 
R = length of rachis in mm. ; and 
6' = number of spikelets. 

This formula expressed in percentage as 

Di % = ^ ^J^^ (Formula 2h) 

R 

was suggested by Neergaard (1887). The latter formula, instead of 
showing how many millimeters of the rachis length on an average are 
shared by each spikelet, as does formula 2a, shows .the expected num- 
ber of spikelets per 100 mm. of rachis length. 

A third and a more logical standard generally used at pres- 
ent consists in determining the average length of the rachis inter- 
node. It is found simply by dividing the length of the rachis by the 
number of internodes (which is always one less than the number of 
spikelets on the same head). This formula may be presented as 

= y (Formula 3a) 

where = density of formula 3a; 

R =■ length of the rachis in mm. ; and 
/ = number of rachis internodes. 

Expressed in percentages, 

Dz % = (Formula 36), 

which shows the average number of rachis internodes found in 100 
mm. of the rachis length. 

Derhtzki (1913) suggested that the true density of a head may be 
better represented by adding i to the percentage of density shown by 
formula 3^ above, thus : 

Di% = ^ ^ or = L>3 + I (Formula 4). 

R 

His formula is a modification of that of Neergaard. The addition 



BOSllNAKIAN : SQUARKIIEAUKDN KSS IN WllKAT. 



of a unit to the luiinber of intornodcs per 100 mni. racliis length is 
based upon the fact that no matter what portion of the rachis is taken 
the number of the spikelets in a ji^iven distance is always one more 
than the number of internodes. luich rachis internode terminates 
with a spikelet, but an additional spikelet without a rachis internode 
is always located at the base of the head just where the basal inter- 
node carrying its spikelet is articulated. It is the presence of the 
basal spikelet which in the calculation of density introduces the addi- 
tion of a unit to the number of internodes which make up an imag- 
inary spike of 100 mm. rachis length. 

Derlitzki's theory appears to be very plausible, as it points out an 
error in Neergaard's formula which for many years was overlooked. 
But as the results obtained from the application of such formulae in 
actual practice are of more interest than the theories themselves, one 
is justified in inquiring just how great the difference is between the 
two formulae in question and,^ further, it is desirable to know how this 
dift'erence compares with the errors which accompany the measuring 
of the material. 

Derlitzki's formula, + be represented as X -|- i by 

/ X 100 

substituting the symbol X for the value — ^ — . Substituting the 

value / + I for the number of spikelets on the head to be measured, 
Neergaard's formula (2b) may be expressed thus,- 
(J + 1) X 100 _ I y. 100 100 _ 100 

R ^~~~^^~R ^^^T' 

2 The value lOO/R is equal to X/I. The product of Neergaard's formula in 
reality is equal to the number of internodes in 100 mm. length plus that portion 

of that coefficient il^^^9) shared by one additional internode. Thus, 
K 

I X 100 
7X100 , R _y , ^ 

For instance, take a head 80 mm. long, with 20 internodes. Then 

I — 20, 

R =80, 

I X 100 2000 
^ = —R- = ^ = 

Density with formula 2b equals 

, 100 , 100 
either X -\ = 25.0 -\ = 25.0 + 1.25 = 26.25, 

R ^ 80 -5 10 

X 25 

or X + — = 25 4 = 25 + 1.25 = 26.25. 

/ 20 



234 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

So we find that the difference between the old {2b) and the cor- 
rected (4) formula is 

Since the variable quantity in this last formula is R, the length of the 
rachis, the difference between the results by these two formulae varies 
inversely with and depends solely upon the length of the rachis pro- 
vided the number of internodes is constant. 

Just how much this difference is for heads of various lengths is 
shown in the following table. Heads with 20 internodes and ranging 
in length from 20 mm. to 160. mm. have been used for illustration. 
The above number of internodes is taken for simplicity and also 
because nearly all wheats, especially the cultivated species, have a 
mean number of internodes fluctuating around that number. 



Table i. — Density of heads of wheat of 20 internodes and varying in length 
from 20 to 160 mm., according to Derlit:2ki's and to Neergaard's formula. 



Length of rachis 
in mm. 
= 


Number of inter- 
nodes per 100 mm. 
= (-V). 


Density after 
Derlitzki's formula 
(A^+i). 


Density after 
Neergaard's formula 


Differences between 
two formulae 

Cf--)- 


20 


100.0 


lOI.O 


105.0 


4.0 


30 


66.7 


67.7 


70.0 


2.3 


40 


50.0 


51.0 


52.5 


1-5 


50 


40.0 


41.0 


42.0 


I.O 


60 


33-3 


34-3 


35-0 


•7 


• 70 


28.7 


29.7 


30.0 


•3 


80 


25.0 


26.0 


26.2 


.2 


90 


22.2 


23.2 


23-3 


.1 


100 


20.0 


21.0 


21.0 


.0 


110 


18.2 


19.2 


19.1 


.1 


120 


16.7 


17.7 


17-5 


.2 


130 


15-4 


16.4 


16. 1 


•3 


140 


14-3 


15-3 


15.0 


•3 


150 


13-3 


14-3 


14.0 


•3 


160 


12.3 


13-3 


12.9 


•4 



The results in the last column of Table i show that the differences 
of the two methods for heads 60 mm. or longer is less than i percent ; 
for heads 5 cm. or shorter the difference varies from i to 4 percent. 

The next questions that arise are : What is the amount of error in 
measuring wheats, and how do the differences between these two 
methods compare with that error? Length measurements in wheat 
are considered satisfactory if measured accurately to the millimeter. 
They cannot be measured to the tenth of a millimeter because : ( i ) 
The end of the rachis is not sharply marked. There is a region about 



BOSHNAKIAN: SQUAKKTIEADKDNKSS in WIIKAT. 



235 



4 mm. long which belongs partly to the spikclet and partly to the 
rachis, and when the spikelet is broken off that region may go with 
the spikelet or may remain attached. A similar region exists at the 
base of the rachis. (2) A slight bending of the rachis may easily 
introduce an error of i to 2 mm., depending on the length of the head. 
(3) The personal er- 
ror in observations is 
too great when dealing 
with such fine meas- 
urements to allow 
measurements of less 
than I mm. Errors 
of 0.5 mm. one w^ay 
or the other are very 
easy to make when 
measurements are 
made with a reason- 
able degree of rapid- 
ity ; errors greater 
than that are also 
made and when one measures a dozen heads on different occasions he 
will find that such errors can not be avoided because the material does 
not lend itself to finer measurements. 

As length is measured to the nearest millimeter, that is, a deviation 
of 0.5 mm. in either direction, the differences in results introduced 



Table 2. — Errors in density determinations of wheat heads introduced by half- 
millimeter errors in measurements of heads of various length. 



Length of heads 
in mm. 


Resuhs of formula 
4 with +0.5 mm. 
error in the length 
measurement. 


Results of formula 
4 with — 0.5 mm. 
error in the length 
measurement. 


Differences of ±0.5 
mm. error. 


Differences of the 
two formulae. 


20 


98. 5 


IOI.5 


3-0 


4.0 


30 


68.8 


66.6 


2.2 


2.4 


40 


51.6 


50.4 


1.2 


1-5 


50 


41.4 


40.6 


.8 


I.O 


60 


34-6 


34-0 


.6 


•7 


70 


29.7 


29-3 


•4 


•4 


80 


26.1 


25-8 


•3 


.2 


90 


23-4 


23.1 


•3 


.1 


100 


21. 1 


20.9 


.2 


.0 


no 


19-3 


19. 1 


.2 


.1 


120 


17.7 


17.6 


.1 


.2 


130 


16.4 


16.3 


.1 


•3 


140 


15-3 


15.2 




•3 


150 


14.4 


14-3 


.1 


•3 


160 


13-5 


13-5 


.0 


•4 




/s no iS ■is f.o 45 io so 6s jy eo es 



Fig. II. Curve showing respective values of 
number of int'ernodes in lOO mm. of rachis length 
and average internode length. 



236 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



by an error of 0.5 mm. will be determined. The same heads having 
20 spikelets each shown in the first column of the preceding table will 
be used and the differences produced by this deviation will be com- 
pared with the differ- 
ences of the two meth- 
ods. The data are 
shown in Table 2. 

It AS seen from the 
last two columns of 
Table 2 that even a 
measurement to the 
nearest millimeter in- 
troduces a difference 
in the results which is 
almost as great as the 
difference between the 
two methods. So far 
as its results in actual 
practice are con- 
cerned, Derlitzki's formula hardly differs from the results obtained 
with Neergaard's formula. 

In what respects do these two methods differ from methods i and 
3? The question may best be answered by comparing the density 
figures obtained by these four methods. Since the last three methods 
represent the density per 100 mm. of rachis length, the first method will 




ZO Z.S 30 3S 4-0 4S SO SS 

Average Ififernoc/e Le/7^/h jn /iMmfi/pr 



7S 8.0 



Fig, 12. Two identical frequency distributions 
plotted for the dense (curve A) and lax (curve B) 
classes. 



be used with the same standard, the formula being 



6'X 100 



(formula 



i&). The writer knows of only one case in which this formula has 
been used. 

Table 3 shows the measurements of six types of heads shown in 
fig. 14. 

The averages of the four methods are here used for comparison, 
but it should not be inferred that these averages represent the figures 
which show the true density. 

The foregoing comparison shows that the least deviation from the 
averages of all four- methods is found in the results obtained by 
formula 3&. The results of formula \h are always below the aver- 
ages, those of formula 2h and 4 always above, whereas those of 
formula 36 fluctuate slightly both above and below. In other words, 
the results by this last formula follow closely the average of the four- 
formulae. These figures show further that with the lax types the 



HOSIINAKTAN : SOUARRII EADKDN KSS IN W 1 1 ICAT. 



difference between the four methods is not so great ljut becomes 
more and more profidunced with the denser forms. 



Table 3. — Mcasurrvicnts of six types of zvlicat heads, with densities calculated 
by various formulae and the deviation by each calculation from the average. 



Data. 



Length of head (L) . . . . . 
Length of rachis {R) .... 

No of spikelets (5) 

No. of internodes (/) . . . 
Percentage of density: 

Formula i& .......... 

Formula 2h 

Formula 36 , 

Formula 4 

Average '. . '. . '. 

Deviations from averages 

Formula ih . 

Formula 26 

Formula 36 

Formula 4 ............ 



Type I. 
Winter 
spelt 
( T. 
spelt a). 


Type 2. 

Amber 
Long- 
berry 

(7'. 7'///- 

^(i > e ), 


Type 3, 
h.arly 
Red 
Chief, 
( 7". 7>ul- 
£<irc). 


Type 4. 
Extra 
Early 
Wind- 
sor, ( T. 
capita- 
tiiin). 


Type 5, 
Little 
Club 
( 7'. com- 
pactuni). 


Type 6, 

f)ale 
Oloria 

( 7'. c<y»i- 
Pacto- 

capita- 
tum). 


Average 
error 
of the 
mean. 


OA 

94 


03 


f\'-i 
07 


<5 1 


40 


42 




87 


78 


61 


75 


41 


35 




20 


20 


18 


20 


19 


20 




19 


19 


17 


19 


18 


19 




21.3 


24.1 


26.9 


24.7 


41-3 


47-6 




23.0 


25.6 


29-5 


26.7 


46.3 


57-1 




21.8 


24-3 


27.8 


25-3 


43-9 


54-3 




22.8 


25-3 


28.8 


26.3 


44.9 


55-3 




22.2 


24.8 


28.2 


25-7 


44.1 


53-6 




-0.9 


-0.7 


-1-3 


— I.O 


-2.8 


-6.0 


±.923 


+ 0.8 


+ 0.8 


+ 1-3 


+ 1.0 


+ 2.2 


+3-5 


±.679 


-0.4 


-0.5 


-0.4 


-0.4 


— 0.2 


+0.7 


±.156 


+0.6 


+ 0.5 


+ 0.6 


+ 0.6 


+ 0.8 


+ 1.7 


±•353 



So far as these results show, formula 36 seems to be the best of 
the four methods. 

It was shown that 
method 3 had two for- 
mulae. The first'( for- 
mula 3a) showed the 
average internode 
length, and the second 
(formula 3&.) "deter- 
mined the density in 
terms of number of 
rachis internodes "ori ^ 
a rachis 100 mm. long. ^ 
Can both be used with 
equal efficiency? 
If, as in the case of 



>~ 6 




9S S>0 as 0a 7S 7ff 6S 69 SS so -^S SS iO 2S 

NumbfT of In/ernodes f>fr MOmtrt. 7?ac/?is /-t^^/h 



Fig. 



13. Values of curves A and B in figure 12 
expressed in terms of number of internodes per 
different standards of rachis length, 

weights, measures, 

etc., with each increase of unit with one formula the figures of 
the other formula -increase or decrease with the same increment, 
either can be used,with equal efficiency. If this be the case the mathe- 



238 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY, 

matical curve showing the corresponding values will take the form of 
a straight line. 

Plotting the figures obtained with both methods, a curve shown in 

figure II is obtained which is the expression of the formula 

It is seen from the curve that the values obtained with the percentage 
basis give more weight to the denser forms, and less weight in the 
same proportion to the laxer types. This fact may be made clearer 
by comparing two normal frequency curves plotted in terms of aver- 
age internode length (fig. 12) and the same values with their respect- 
ive frequencies in terms of number of internodes per 100 mm. rachis 
length (fig. 13). The frequencies and corresponding classes and 
results of both standards are shown in Table 4. 



Table 4. — Frequencies in classes of average internode length and classes of 
number of internodes in 100 mm. in curves A and B {fig. 12) and 
curves A' and B' {fig. 13). 



Curve A (fig. 12). 


Curve A' (fig. 13). 




Curve B (fig. 12). 


Curve i?' (fig. 13). 




Classes of 
average 
internode 
length in 
mm. 


Difference 
between 
classes, 
mm. 


Classes of 
number of 
internodes 
in 100 mm. « 


Difference 
between 
classes, 
mm. 


Fre- 
quen- 
cies. 


Classes of 
average 
internode 
length in 
mm. 


Difference 
between 
classes, 
mm. 


Classes of 
number of 
internodes 
in 100 mm.* 


Difference 
between 
classes, 
mm. 


Fre- 
quen- 
cies. 


1- 5 
2.0 

2- 5 

3- 

3- 5 
4.0 

4- 5 


0.5 
0.5 
0.5 
0.5 
0.5 
0.5 


66.6 
50.0 
40.0 
33-3 
28.5 
25.0 
22.2 


16.6 
10. 
6.7 
4.8 

3-5 
2.8 


I 

6 

^5 
20 

15 
6 
I 


4- 5 

5- 
5-5 
6.0 

6.5 
7.0 

7-5 


0.5 
0.5 
0.5 
0.5 
0.5 
0.5 


22.2 
20.0 
18.2 
16.6 

15-4 
14-3 
13-3 


2.2 
1.8 
1.6 
1.2 
I.I 
I.O 


I 
6 

15 
20 

15 
6 
I 


M =3.o±.052 
a =.612 ±.036 


M =34.82 ±.682 
a =8.093 ±.482 




M =6.o±.052 
<r =.612 ±.036 


M = 16.92 ±.144 
0- = i.703±.i02 





* Equivalents of classes of curve A (fig. 12). 

* Equivalents of classes of curve B (fig. 12). 



The figures obtained above point out four difTerences in the use 
of the two formulae, 3a and : 

1. A normal frequency curve with formula 3a gives a skew curve 
when plotted in terms of number of internodes per 100 mm., the 
skewness being toward the more lax classes. This is well shown by 
curves A and A' plotted in figs. 2 and 3. Curve A (fig. 12) is a 
symmetrical curve where the mode 3.0 coincides with the median, the 
center of the range. The median of curve A' (fig. 13) is at 44.4 but 
the mode is at 33.3, or ii.i internodes towards the laxer classes. 



BOSIINAKIAN: SQUAUIJIKADKDNKSS in WIIKAT. 



239 



2. On account of this skcvvncss the mean density oljtained with 
formula is always greater than the corresponding value of the 
mean obtained with formula 3a. When the mean internode length of 
the frequency distribution represented by curve A is 3 mm. its corre- 
sponding value with formula 3^ should be 33.33 internodes per 100 
mm. Calculation, however, shows 34.82 internodes, an increase of 
1.49 internodes. With curve B', where the range on the percentage 
basis is very small, the difference is not so great. The equivalent of 
6 mm., the mean of curve B, is 16.66 internodes per 100 mm. The 
calculated mean is 16.92, wdiich shows a difference of +0.26 inter- 
node. 

3. Frequency curves of dense and lax types found by formula 3a 
(fig. 12) show* when expressed with formula 3Z7, an appreciably 
greater range of variation with denser wheats (fig. 13, curve A) and 
a much smaller range among laxer forms (fig. 13, curve B). While 
the ranges of variation of the two curves with formula 3a are iden- 
tical, these same curves plotted in terms of number of internodes per 
100 mm. will show ranges of 44.4 and 8.9 internodes for curves A 
and B respectively. This gives the impression that the range of curve 
A is five times greater than that of curve B, which of course is 
erroneous. 

4. Consequently, the standard deviation of formula 36 tends to 
increase and decrease considerably as the figures of the classes 
increase or decrease. The standard deviations of curves A and B 
(formula 3a) were found to be identical, that is, 0.612 ± .036 in both 
cases. As calculated for the other formula (3^) the standard devia- 
tions are 8.093 zh .482 and 1.703 zh .102 respectively, showing an 
appreciable difference of 7.390 ± .492 internodes. 



Table 5. — Variation and range of density in Dale Gloria and Amber Longberry 

wheats. 





Dale Gloria. 


Amber Longberry. 


Ratio of ranges. 


Formula 3a: 








Variation 


1.2 to 2.3 mm. 


3.0 to 5.4 mm. 




Range 


I.I mm. 


2.4 mm. 


1:2.18 


Formula 3b: 








Variation 


83-3 to 43.5 percent 


33.3 to 18.5 percent 




Range 


39.8 percent 


14.8 percent 


1:0.37 



In nature, the range of variation and the standard deviation of 
length characters, such as those of internodes, spikes, and culms in 
wheat, decreases as the length of the character in question decreases. 
For instance, in comparing the range of variation of density of Dale 



240 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Gloria (type 6, fig. 14) and of Amber Longberry (type 2) with the 
two formulae the results shown in Table 5 are obtained. 

It is seen from the above actual case that the range of density of 
Amber Longberry is over twice that of Dale Gloria (formula 3a). 
On the other hand, as found on the percentage basis (formula 3&), 
the fact seems to be just the reverse. The figures of the percentage 
formula show that Dale Gloria has a range over two and one-half 
times that of Amber Longberry. Is this latter claim justified? Cer- 
tainly not. It can only be said that Dale Gloria varies from 83.3 to 
43.5 internodes per 100 mm. of rachis length and Amber Longberry 
from 33.3 to 18.5. One must bear in mind that a difference of 5 
internodes between classes 15 and 20, for instance, is entirely differ- 
ent from the same difference between classes 85 and 90. While a 
difference of 5 internodes between classes 85 and 90 makes a differ- 
ence of 0.07 mm. on the average internode length, the same difference 
of 5 internodes between classes 10-15 produces a difference of 1.66 
mm., a figure 14.4 times greater. 

Here one confronts another difficulty. What does this standard 
deviation of a distribution such as that of curve A' (fig. 13) show? 
Briefly, it shows hardly anything, because the standard deviation 
which is meant to show some form of departure above and below the 
mean can not be applied to a distribution where the density values of 
classes, as was shown above, are different. Referring to curve A' 
(fig. 13) the calculated standard deviation was found to be 8.093 
internodes per 100 mm. rachis length. As 8.093 internodes below the 
mean have a density value several times greater than the value of the 
same number of internodes above the mean, our standard deviation 
thus calculated fails to express the type of deviation which the coeffi- 
cient under consideration is expected to show. 

It must not be concluded, however, from the above discussion that 
expression of density in terms of number of internodes per 100 mm. 
rachis length is itself absurd. On the contrary, it is mathematically 
correct, but it is misleading and does not lend itself easily to the appli- 
cation of statistical methods which are commonly in use. 

What was said above for formula 3& applies also to formulae ib, 
2b, and 4, because the principle upon which they work is essentially 
the same. 

Two other formulae were mentioned which did not express density 
on the percentage basis. They were formula la, where the density 
was found by dividing the number of spikelets into the total length of 
the head, and formula 2a, where the number of spikelets was divided 



boshnakian: sguARi<:iiKAi)i:i)Ni:ss in vviikat. 241 

into the lenj^th of the rachis. It was already said at the bej,nnning of 
this pai)er that density found with the first formula (la) depended 
largely upon the length of the terminal spikelet, the error of which 
would increase with the shortness of the head. As to the second 
formula, it can neither be used in place of formula 3a nor preferred 
over it because the standard taken is not logical. There is no reason 
why density should be expressed by dividing the number of spikclets 
into the length of the rachis, because the rachis itself is not made of 
spikelets but of rachis internodes ; furthermore, the number of spike- 
lets does not correspond to the number of rachis internodes. 

In conclusion it may be said that of all the formulae given thus 
far the average internode length represents the best method for 
determining density, as density is dependent directly upon the length 
of the rachis and the number of its units, the internodes, of which it 
is composed. 

2. The Coefficient of the Squarehead Forms.^ 

The average rachis internode-length which is the best index of 
density suggested has certain limits. It can not bring out the differ- 
ence between a squarehead and a nonsquarehead. A comparison of 
types 2, 3, 4, 5, and 6 is sufficient to show this (Table 6). 

Table 6. — Comparison of internode lengths of several types of wheat. 



Type number and species. 



Measurement. 


2. T. vul- 


3. T. vul- 






6. T. coni- 




4. T. capi- 


5. T. com- 


pacto-capi- 




gar e. 


gar e. 


tation. 


pactum. 


tatian. 


Rachis length, mm 


78 


61 


75 


. 41 


35 


No. of internodes 


19 


17 


19 


18 


19 


Av. internode length, mm. . 


4.10 


3-59 


3-95 


2.28 


1.84 



We know from the diagrams in fig. 14 that types 4 and 6 are 
squarehead forms, but this index of density fails to reveal it. For 
instance, the density of type 4, a typical squarehead, is 3.95 ; it is 
slightly below the internode length of type 2 and above that of type 
3, both vulgare form's. There is no marked diflference between the 
internode length of the squarehead and that of the vulgare forms. 

The same inability of distinguishing the squarehead form from a 
nonsquarehead form is found when the density of the same square- 
head (type 4) is compared with that of the compactum form (type 

3 A paper presented at the annual meeting of the American Genetic Associa- 
tion, Berkeley, Cal., August 5, 191 5. 



242 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



5). The difference between the squarehead and the other forms is 
not merely a matter of density. Therefore, the use of any formula 
showing mere density is of no value when the object is to bring out 
the character of squareheadedness. 

The difference between this form and the vulgar e form particularly 
is the compactness of the terminal portion of the spike, which is the 
direct result of the shortening of the internodes in that region. On 
account of this shortening, the terminal spikelets mechanically diverge 
(see diagram) in order to have more room for development. As the 
tip of the .diverging spikelet is farther from the rachis than that of 
the spikelet lying close to the rachis (type i), the width of the head 
in the region of shorter internodes is greater than that of the basal 
portion. This widening of the tip has suggested the name of club." 
On account of the widening of the tip and some other factors which 



side it gives the impression that the head is square in cross-section, 
although in reality it is more or less triangular. On account 
of this popular belief heads showing this character are called 
" squareheads." Because of similar differences in internode length 
one of the species of barley is popularly called " four-rowed " as dis- 
tinguished from the " six-rowed," although there is no such a thing 
as " four-rowed " barley, as a close examination will readily show. 

Let it be pointed out also in this connection that squarehead and 
nonsquarehead forms exist among the denser types. Types 5 and 6 
respectively show this difference. These two forms are now classi- 
fied as Triticum compactum, distinguishing them from T. capitatum, 



90. 
W_ 

40 _ 
30. 

JO - 
0_ 



Fig. 14. Diagrams showing the arrangement of 
the spikelets on the rachis in several types of 
wheat. 





need not be mentioned 
in this paper, the head 
makes a twisting move- 
ment a day or two be- 
fore and also during 
the period of heading; 
this twisting draws the 
terminal spikelets to- 
ward one side, exposing 
to view one side of the 
rachis. Where the 
spikelets are drawn 
away the surface ap- 
pears somewhat flat, 
and when the head is 
viewed from this flat 



150SHNAKIAN : SQUAUKI I KADICDN lOSS IN WIIICAT. 



the squarehead. As the apparent difference and the genetic behavior 
between ty[)es 5 and 6 are as marked as those of types 2 and 4, a (hs- 
tinction needs to be made. For tliis reason it is suggested in this 
j)apcr that T. coinpacl iini be used for compact wheats which lack 
s(|uareheadedncss and T. compaclo-capitalum for compact, square- 
headed forms. The fact that uniform and squarehead forms exist 
in dense as well as in lax types needs be considered in making classi- 
fication of cultivated wheats. 

Having discussed the differences between the squarehead and uni- 
formly dense types, the question arises as to how to express this 
dift'ercnce and all its gradations in terms of a coefficient which may 
be adopted in statistical work. 

Since squareheadedness is the result of the shortening of the 
terminal internodes, a number of methods were tried to bring out 
this difference. It was found that the ratio between the number of 
internodes in the middle third of the rachis and the number of inter- 
nodes in the upper third would best express the degree of square- 
headedness. This coefficient of squareheadedness, represented by the 
symbol " Sq,'' is found by the formula 

12 

where Sq = coefficient of squareheadedness ; 

= number of internodes in the terminal third of the rachis; and 
/j = number of internodes in the middle third of the rachis. 

This formula represents in reahty the relative density of these two 
regions. This is elucidated in the following formula: 

R 

Sq = -7- ; R = length of rachis. 

R 
3 

Cancelling R/^ in both the numerator and the denominator the 
simpHfied formula already given is obtained. The results shown in 
Table 7 are obtained from the use of this formula. 

The new coefficient, it is seen, has brought out very markedly the 
difference between the squareheads and other forms. It shows, for 
instance, in the case of type 3 that there is not much difference 
between the densities of the two regions in question, while in type 4 
the terminal portion is 1.55 times denser than the middle third. It 
is also of special interest to notice that T. compactum (type 5), 



244 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



although dense, was not affected at all. This would naturally be 
expected, for the rachis internodes of the latter are short but more 
or less uniform. 

Table 7. — Coefficients of squareheadedness in several types of wheat, as deter- 
mined by the formula, Sq = Y- 



Type number and species. 



Measurement. 


2. T. vul- 
gar e. 


3. T. vul- 
gar e. 


4. T. capi- 
iatum. 


5. T. com- 
pactum. 


6. T. com- 
pacto-capi- 
tatmn. 


No. of internodes in upper third of 












rachis 


6.2 


5-7 


8.4 


5.8 


8.0 


No. of internodes in middle third of 












rachis 


5-6 


5-3 


5-2 


5.6 


5.1 


Coefificient of squareheadedness (Sq.) . . 


1. 10 


1.07 


1-55 


1.03 


1.53 



In describing a compact form the use of density as found in terms 
of average internode length should not be neglected altogether, for 
the new coefficient is meant to represent degrees of squareheadedness 
only. If a clear idea of the shape and length of the head is desired 
the density should accompany this coefficient. Thus, types 4 and 5 
are represented as 

Type 4 6'(7=:i.S5, ^==3.95, 

Type 5 • Sq — 1.03, D — 2.28. 

These figures show that type 4 is a long, open squarehead, and type 
5 a short and dense wheat of uniform internode length.* 

When the material to be measured is grown in the greenhouse or 
has certain abnormalities, such as sterile or rudimentary terminal 
internodes or abnormally long basal internodes, it will be found more 
satisfactory to express the density in terms of the average internode 
length of the middle third of the rachis. Where the head develops 
normally, there is no reason why the average internode length of the 
head should not be used. 

3. The Instrument for Measuring Squareheadedness. 

When the density or squarehead coefficient of a large number of 
heads is to be determined, the use of an instrument which will do the 
work rapidly and accurately becomes necessary. An instrument 
devised by the writer for this purpose is shown in Plate 7, and a 

4 Experience has shown that all gradations exist between the squarehead and 
vulgare forms, and the forms which by sight might be classified as squareheads 
have a coefficient almost always above 1.33, 



246 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

working plan, slightly modified, in figure 15. The instrument consists 
of a rectangular wooden base A (fig. 15, top view) over which by- 
means of wooden blocks, c and c', two arms equal in length, p and 
are supported. The arms are held parallel to each other with a cross 
bar m, the distance between the centers of the rivets x and x' being 
exactly equal to the distance between centers at x" and x'" where the 
arms are pivoted. Along the sides of the arms which face each other, 
brass plates, g and g' , are fastened. These brass plates carry five 
pairs of holes large enough for a thread to pass through ; three of 
these pairs, r, r' and r" , are spaced 1.66 cm. from each other, as 
shown in the drawing (side view), so that the last pair, is 5 cm. 
from the stationary centers x" or x'" ; the other two pairs, .y and s\ 
are 5 cm. apart from each other. These pairs of holes are perpen- 
dicular to the long side of the plate and therefore parallel to each 
other. Through the corresponding holes on brass plates g and g^ 
which face each other pass waxed silk threads each pair of which is 
held in tension by coil springs, h, h' , etc. The details of this arrange- 
ment are shown in the lower right-hand corner of figure 15. 

The first set of threads, s'-s' , s-s, and r"-r" , divide the distance 
between the cross thread s-s' and the imaginary base line 0-0' into 
three equal parts by parallel planes of vision which pass through each 
pair of threads. The distance between imaginary line 0-0' and r"-r" 
is also divided into three equal parts by a second set of threads, r'-r' ^ 
and r-r. When the arms of an accurately constructed instrument are 
in normal position, as shown in full lines in the drawing, the threads 
s' , s, r", r', r, should register on the metric scale 15.00, 10.00, 5.00, 2.33,. 
and 1.66 cm., respectively. By moving the arms towards the right 
to the position shown in dotted lines, for instance, the movable parts 
assume the form of a parallelogram. As the distance between s'-s' 
and 0-0' decreases, the distance between each pair of threads 
decreases also in the same ratio ; the threads of each set, therefore, 
remain always equidistant and parallel to each other. These are the 
principles upon which this instrument is constructed. 

The manipulation of the instrument just described is extremely 
simple. If the length of the rachis is less than 5 cm. the lower set of 
threads (r"-r', r'-r', r-r) is used for dividing the head into three 
equal parts, otherwise the upper set {s'-s' , s-s, r"-r") is used. The 
culm of the head to be measured is held firmly with a clasp provided 
for this purpose, and the base of the rachis brought in line with the 
imaginary line 0-0' (fig. 15). In the drawing this line is repre- 
sented by the upper edge of the block q and in the illustration (Plate 



HOSTl NAKTAN : SOU AK i: 1 1 IvMJiCDN KSS IN WIIICAT. 



7) by a while object i)laccd in the same position. Now the movable 
parts are swuiii^' so that cross thread s'-s' is in hne with the ti]) of the 
rachis (fij;-. 15). The spike is thns (hvided into three eqnal parts. 
As the formula for the squarehead form calls for the counts of the 
number of internodes in the U])per and middle third of the rachis 
only, these data only need to be taken. In the case of the head in the 
illustration, the upper third contains 13 rachis internodes and the 
middle third 6.4 ; the coefficient of the squarehead form is therefore 
13-^6.4 = 2.03. 

Such data are usually accompanied with the average internode 
length. Therefore the length of the rachis registered by the upper 
cross thread on the metric rule is also recorded and the count of the 
total internodes made later, or if the material is such that the average 
internode length does not give a satisfactory index of the compact- 
ness, as in the case of some deformities of the basal rachis internodes, 
then the third of the length of the head as registered by cross-thread 
r"-r" is recorded. In this case it is not necessary to count internodes, 
as the count of the internodes at the middle third has already been 
recorded. According to the formula this latter figure is divided into 
the third of the length of the rachis and a figure is obtained which 
shows the average internode length of the middle third of the rachis. 

This instrument, then, performs simultaneously three operations : 
(i) It divides the rachis into three equal parts, (2) it registers the 
length of the rachis, and (3) it registers the third of the length of 
the rachis. 

Literature Cited. 

Neergaard, Th. von. 

1887. Normalsystem for bedomande af axets morfologiska sammansatt- 
ning hos vera sadesslag. All. svenska Utsodesforenningens arbe- 
rattelse for ar 1887, p. 37. Cit. Kondo, Landw. Jahrb. 45: 711-817. 

Derlitzki, G. 

1913. Beitrage zur Systematik des Roggens durch Untersuchungen iiber 
den Ahrenbau. Landw. Jahrb. 44 : 353-407. 



248 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



THE MOISTURE CONTENT OF HEATING WHEAT.^ 

C. H. Bailey. 

During the latter part of July and first half of August, 1916, an 
unusually large proportion of the wheat sampled by the Minnesota 
Grain Inspection Department was found to be in a heating condition. 
Several hundred cars of heating spring wheat were received at Min- 
neapolis, in addition to large quantities of other cereals. The con- 
dition of this grain varied from incipient heating to a badly bin- 
burned or mahogany-colored character. 

The weather from the middle of July to the middle of August, 
1 91 6, was characterized by unusually high temperatures. The mean 
of the maximum daily temperatures for July, 1916, at Minneapolis 
was 88.3° F., while the same record for July, 1915, was 75.5°. Simi- 
larly, the mean of the maximum daily temperatures for August, 1916, 
was 81.7° F., while the same month of the preceding year it was 75.0°. 
This hot weather induced heating in grain which would have remained 
sound under normal conditions. It is known that a high tempera- 
ture accelerates respiration in grain, which respiration, if sufficiently 
rapid, results in the phenomenon known as heating, or bin-burning. 
If the enzymic changes known in the aggregate as respiration follow 
the usual rules with regard to the relative acceleration of enzyme 
activity resulting from an increase in temperature, an increase of 
from two and a half to three times in the rate of respiration should 
result from a rise of 18° F. (10° C.) in temperature. This would 
effect a corresponding increase in the amount of heat developed as 
the result of the biological combustion or oxidation which constitutes 
an important part of respiration. Moreover, a smaller proportion 
of the heat developed is radiated or conducted from heating grain 
into a surrounding hot atmosphere than into cool air. Accordingly, 
hot weather accelerates heating in grain in two ways, biochemically 
through the effect on the rate of enzyme action, and physically by 
reducing the rate of conduction of heat into the atmosphere. 

Rumors became current in Minneapolis that dry " wheat was 
heating as readily as the damp grain. In order to ascertain the accu- 
racy of such statements and to determine the actual moisture content 

^ Contribution from the Minnesota Grain Inspection Department Laboratory. 
Received for publication January 27, 1917. 



BAILKV : MOISTURK CONTl-.NT OK IIKATINC WHEAT. 249 

of heating grain, the writer personally examined and analyzed the 
samples drawn from a large number of cars. Mr. Albert Nelson was 
detailed to assist in the investigation and his painstaking, accurate 
work contributed materially to its progress. 

For the purposes of our work it was found most convenient to 
visit the sidings of certain of the grain hospitals engaged in condi- 
tioning the hot wheat, in order to gain access to the largest possible 
number of cars each day. During the 14 days from August 3 to 16 
about 70 cars of various types of heating wheat were examined in 
this manner, in addition to a number of lots of other grain. 

These studies served to emphasize the necessity of placing the sam- 
ples in tight containers as soon as drawm from the bulk. In hot 
weather a 3-pound sample in a cloth sack will lose several tenths of 
I percent of moisture while it is being transported from the yards to 
the office, and if left in the sack over night the loss of moisture may 
amount to several percent. A failure properly to prevent evaporation 
from this grain was probably the cause of the erroneous ideas con- 
cerning the moisture content of some of the heating w^heat. All the 
samples taken in this work were placed immediately in 4-ounce glass 
bottles which \vere tightly stoppered. These were not opened until 
a portion was weighed into a covered aluminum drying capsule. The 
moisture was determined by drying the unground kernels in an air 
oven at a temperature of 98° C. 

In tabulating the data the samples were arranged in order of their 
moisture content, and are grouped with a range of five-tenths of i 
percent of moisture in each group. Such information as was fur- 
nished by the railroad companies concerning the source of the grain 
and the date on which it was loaded into cars was incorporated in the 
table. In a few instances no such record could be obtained. No 
information concerning the previous history of the grain before it 
was loaded into the cars could be gotten. It is probable, however, 
that in many instances the grain was in a heating condition before it 
was shipped. 

In discussing these data with members of the grain inspection 
service the question was raised as to how much moisture may have 
been lost by evaporation from the heating mass. Undoubtedly some 
loss had occurred in this manner from many of the lots examined. 
The amount of loss depends in large part upon the temperature which 
it had reached and the length of exposure. With the data available 
any estimate of the loss would be pure speculation, and no attempts 
were made in that direction. One thing appears evident, however; 
the moisture content of sound, plump spring wheat must be above 



250 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Table i. — Moisture content and other data relating to heating wheat sampled 
at Minneapolis, Minn., August 3-16, 1916. 









T TTCC 


THAN 


LZ^ iril^KCliiM X JMUioXUivE.. 








Lab. 
No. 


Moisture. 


Dock- 
age per 
bushel. 


Weight 

per 
bushel. 


Weight 
per 
1,000 
kernels. 


Consigned from — 


Date 
loaded. 


Date 
sampled. 


Days 
in the 
car. 


119a 
46a 


Percent. 
13.50 
13.78 


Lbs. 
2 

2h 


Lbs. 
57 
55I 


Grams. 
26.64 
23.16 


Langdon, N. Dak. 
Gibbon, Minn. 


July 19 
July 20 


Aug. 12 

Aug. 3 


24 
14 


14.01 PERCENT TO I4.5O PERCENT MOISTURE. 


142a 
850 
94a 
88a 
84a 
43 a 
89a 
99a 
72a 

143a 

Ii6a 
75a 

115a 


14.04 
14. II 
14.19 
14.21 
14.27 
14.34 
14.34 
14.36 
14.43 
14.43 
14.44 

14.45 
14.49 


2 

3 

4l 

2h 

I 

2 

l| 
l| 
I 

2h 
ih 
2\ 
I* 


52 
53 

54 
50 

55 
59 
56 
55 

54 


19.56 
24.76 
21.24 
21.60 
17.16 
27.76 
27.96 
27.48 
23.62 
24.52 
32.80 
24.20 
22.60 


Eureka, S. Dak. 
Pierpont, S. Dak. 
Pioneer Steel Elevator 
Tracy, Minn. 
Victoria Elevator 
Perella, N. Dak. 
Brittin, N. Dak. 
Dunn Center, N. Dak. 
Timmer, N. Dak. 
Longwood, N. Dak. 
Mowbray, N. Dak. 
Killdeer, N. Dak. 
Marfield Elevator 


July 28 
July 29 
July II 
Aug. I 
July 29 

July 26 
June 27 
July 20 
Aug. 4 
Aug. 2 
July ?6 
July 28 


Aug. 14 
Aug. 8 
Aug. 9 
Aug. 8 
Aug. 8 
Aug. 3 
Aug. 8 
Aug. 9 
Aug. 7 
Aug. 14 
Aug. II 
Aug. 7 
Aug. II 


17 
10 
29 

7 
10 

13 
43 
18 
10 

9 
12 

14 


14.51 PERCENT TO I5.OO PERCENT MOISTURE. 


114a 

73a 

141a 
96a 
50a 

146a 
49a 

123a 
97a 

ii8a 
74a 
77a 

124a 
80a 
44a 

I2ia 


14.52 
14.56 
14.56 
14.60 
14.66 
14.68 
14.70 
14.77 
14.81 
14.83 
14.86 
14.88 
14.94 
14.97 
14.98 
15.00 


l| 
I 

2\ 

1 
2 

3 

1 
2 

1 
2 
1 
2 

2 
2 


56^ 
55 

54 

54 

53i 

55 

56 

56 
57 

• 56I 


23.44 
24.62 
28.68 
29.52 
17.58 
27.56 
20.02 
25.16 
23.00 
20.96 
25.38 
18.90 
28.24 
27.84 
25.72 
24.64 


Doland, S. Dak. 
Calumet Elevator 
No record 
Demming, N. Dak. 
No record 
Marfield Elevator 
Eureka, S. Dak. 
No record 
Dickinson Elevator 
Wilton, N. Dak. 
Kansas City, Mo. 
Platte, S. Dak. 
Kempton, N. Dak. 
Lawton, N. Dak. 
Talmo, N. Dak. 
Bixby, Minn. 


Aug. 2 
July 21 

Aug. I 

July 29 
July 17 

Aug. 2 
July 23 
July 27 
June 30 
Aug. 5 
July 31 
July 15 
Aug. 8 


Aug. II 
Aug 7 
Aug. 14 
Aug. 7 
Aug. 3 
Aug. 15 
Aug. 3 
Aug. 12 
Aug. 9 
Aug. II 
Aug. 7 
Aug. 7 
Aug. 12 
Aug. 7 
Aug. 3 
Aug. 12 


9 
17 

6 

17 
17 

7 
19 
II 
38 

7 

7 
19 

4 


15.01 PERCENT TO I5.5O PERCENT MOISTURE. 


92a 
125a 
loia 

51a 
144a 

71a 
149a 
io6a 

79a 


15.02 
15.06 
15. II 
15.15 
15.15 
15.18 
15.30 
15.34 
15.40 


I 
4 
I 

6^ 
I 

2 

2\ 
I 

3 


55 

55 

57 

47 

52I 

58 

59 
56 


28.52 
24.80 
29.56 
14.52 
22.44 
30.12 
26.08 
28.72 
26.74 


Wall. S. Dak. 
Killdeer, N. Dak. 
Wales, N. Dak. 

Glenville, Minn. 
McHenry, N. Dak. 
Phelps. N. Dak. 
Marfield Elevator 
Lawton, N. Dak. 


Aug. 2 
July 26 
July 28 

Aug. 8 
July 26 
Aug. 3 
July 26 
July 26 


Aug. 9 
Aug. 12 
Aug. 9 
Aug. 3 
Aug. 15 
Aug. 7 
Aug. 16 
Aug. 10 
Aug. 7 


7 
17 
12 

7 
12 

13 
15 
12 



bailey: moisturk (ontknt of jii-:atin(; wjieat. 



251 



Over 15.51 percent moisture. 



Lab. 
No. 



Moisture. 


Dock- 
age per 
bushel. 


Weight 

per 
bushel. 


Weight 
per 
1,000 
kernels. 


Consigned from — 


Date 
loaded . 


Date 
sampled. 


Days 
in tbe 
car. 


Percent. 


L.OS. 


L.OS. 


Grams. 










15-52 


1 
2 


54 






A 11 T 


Aug. 10 



v 


1554 






30.36 


Gwinner, N. Dak. 


Aug. 3 


Aug. 10 


7 


15-56 






19.24 


No record 




Aug. 3 




15-59 


1 

"2 


55 


22.92 


Eureka, S. Dak. 


Aug. 4 


Aug. 12 


8 


15-68 






27.92 






Aug. 10 




15-66 


3 


52 


20.88 


Ashley, N. Dak. 


Aug. 10 


Aug. 16 


6 


15-70 


5 


56 


30.28 


No record 




Aug. 7 




i5-«7 


h 




18.92 


Westfield, Iowa 


Aug. 7 


Aug. 15 


8 


16.15 


2h 


56 


25.84 


Coteau, N. Dak. 


July 28 


Aug. 12 


15 



the normal (about 13.75 percent) before heating ensues, even under 
the extreme conditions of the hot summer weather of 191 6, as there 
is no reason to believe that there was any material increase in the 
moisture content after the grain started to heat. Two samples were 
taken from heating wheat which contained less than 14 percent of 
moisture but neither of these were normal grain. Sample 119a was 
frosted, while 46a was shriveled. It will be observed that all the 
heating samples containing less than 14.3 percent of moisture were 
shriveled, with a low weight per bushel, indicating a greater tendency 
on the part of such grain to heat. Since all of the normally plump 
spring wheat that heated contained over 14.3 percent of moisture, and 
there is reason to believe that the moisture content before heating 
commenced was higher, the writer concludes that sound, plump, hard 
wheat containing less than 14.5 percent of moisture will keep without 
heating in storage in a temperate climate. A lower moisture limit 
must be employed in storing shriveled or frosted wheat, and possibly 
with sound, plump wheat in a tropical climate. 

Minnesota Grain Inspection Dept. Laboratory, 
Minneapolis, Minn. 



252 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



AGRONOMIC AFFAIRS. 



NEW BOOKS. 

• 

Organic Agricultural Chemistry (The Chemistry of Plants and Ani- 
mals). By Joseph Scudder Chamberlain, Professor of Agricultural 
Chemistry, Massachusetts Agricultural College. The Macmillan 
Co., New York, 1916. Pages xvii plus 319. 

The subject matter of this book is divided into three sections en- 
titled, respectively. Systematic, Physiological, and Crops, Foods and 
Feeding. 

The first section includes what is the first satisfactory survey which 
is known to the reviewer of the general and fundamental principles 
of organic chemistry from the viewpoint of the student or instructor 
who desires to make use of the subject for the study of plant and 
animal life rather than for advanced study of organic compounds and 
relationships of scientific interest. The presentation, though neces- 
sarily brief and including only those types of carbon compounds which 
are involved in nutritional or physiological processes, is carefully 
worked out and is sufficient for most collegiate students of plant, 
animal, or human nutrition. The omission of any discussion of the 
closed-ring compounds is a sacrifice to brevity which will make neces- 
sary some supplemental instruction, if the physiological function of 
certain important units of the proteins is to be understood. The post- 
ponement of the discussion of the ''unsaturated compounds" to the 
last pages of the section makes the consideration of the chemistry of 
the fats and of certain of the alcohols and mixed compounds rather 
inadequate, and an earlier consideration of these would improve the 
whole section. But these are only minor faults in what is otherwise a 
very satisfactory arrangement and choice of subject matter. 

The second section deals, in successive chapters, with enzymes and 
enzymatic action ; the living cell and its food ; animal food and nutri- 
tion; milk, blood, and urine; and plant physiology. The departure 
from the usual custom, namely, the consideration of the nutrition of 
animals before the study of the biochemistry of plants, is considered 
by the author to emphasize " the real difference between these two 
forms while retaining the idea of fundamental similarity." The re- 



AGRONOMIC M'l-AIRS. 



viewer prefers the custoniary arrangement of studying first the syn- 
thesis of organic matter from its elements by plants, followed by the 
more complex syntheses and reactions which constitute animal metab- 
olism, but recognizes as a commendable feature of the innovation 
that the unity rather than the dissimilarity of these two phases of bio- 
chemistry is emphasized. The subject matter of this section, though 
brief and simply presented, is admirably selected and will prove an 
excellent preparation for further studies of practical nutritional 
problems. 

Section III contains two chapters dealing with the occurrence and 
uses of important constituents in agricultural plants, in which the 
composition of the more common carbohydrates, fats, lipoids, and 
proteins, and their industrial and nutritional uses are briefly discussed. 
A third and final chapter contains a very brief but carefully prepared 
discussion of the quantitative relations of food consumption to energy 
production in the animal body. The subject matter of the entire sec- 
tion carefully avoids controversial discussions and presents in an ad- 
mirable way a clear and satisfactory survey of the biochemical prin- 
ciples involved. 

The book contains a few typographical errors, which will undoubt- 
edly be corrected in later editions, but is commendably free from mis- 
statements or comfusing personal idiosyncrasies of expression or form. 
It is bound uniformly with the familiar Macmillan series, and the 
absence of illustrations has permitted it to be printed on the desirable 
rough surface book paper. The reviewer predicts a very general use 
of the book as a textbook for college students who are preparing for 
further study of animal or human nutrition, and regards it as a thor- 
oughly satisfactory text for that phase of agricultural chemistry which 
may be designated as biochemistry " to distinguish it from the study 
of soils and fertilizers, which is now commonly taught as a separate 
subject and from an entirely different viewpoint. — R. W. Thatcher. 

NO SUMMER ISSUES OF THE JOURNAL. 

As announced earlier in the year, the Journal of the American 
Society of Agronomy will not be published during June, July, and 
August. The next issue will appear about September 15. While 
monthly issuance is desirable, our present income does not justify it. 
Those who feel that the publication of twelve numbers during the 
year is desirable can help in making such publication possible by ob- 
taining new members for the Society. We need 100 more members 
this year. 



2 54 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



MEMBERSHIP CHANGES. 



The membership of the Society, as reported in the April issue, was 
663. Since that time 13 new members have been added and one has 
been reinstated. In the same period 4 members have resigned, i has 
died, and 29 have lapsed for nonpayment of dues for 1916. The net 
loss in membership is 20, making a total membership at this time of 
643. The names and addresses of the new and reinstated members, 
with the names of those who have resigned and lapsed and of the 
deceased member, with such changes of address as have come to the 
notice of the Secretary, are as follows : 



Booth, V. J., 318 West Street, Stillwater, Okla. 
Darst, W. H., Dept. of Agronomy, State College, Pa. 
Dickson, R. E., Substation No. 7, Spur, Texas. 
Freeman, Ray, Box 255, Fort Worth, Tex. 

Hill, C. Edwin, Eastern Oregon Dry-Farming Substation, Moro, Oreg. 

Hill, Pope R., Box 625, Athens, Ga. 

Milton, R. H., Clarksville, Tenn. 

Robinson, R. B., Box 272, Stillwater, Okla. 

Taggart, J. G., School of Agriculture, Olds, Alta., Canada. 

ToRGERSON, E. F., 654 Agricultural Bldg., University of Illinois, Urbana, 111. 

Trout, C. E., Bur. Plant Indus., U. S. Dept. Agr., Washington, D. C. 

Van Evera, R., 16 Mendota Court, Madison, Wis. 

Van Nuis, C. S., College Farm, New Brunswick, N. J. 



New Members. 



Members Reinstated. 



Starr, S. H., College of Agriculture, Athens, Ga. 



Members Resigned. 



Hendrick, H. B., 



Morse, Fred W., 
Scales, Freeman M. 



Peacock, Walter M., 



Members Lapsed. 



Bassett, L. B., 
Benton, T. H., 
Brown, C. B., 
Cassel, Chas. E., 
Chapman, Jas. E., 
Clothier, R. W., 
CuRREY, Hiram M., 
Damon, S. C, 
Gilbert, Arthur W., 
Hershberger, Jos. P., 



Holland, Robt. E., 
Hughes, H. D., 
Khankhoje, p. S., 



Kinney, H. B., 
Lechner, H. J., 



LuMBRicK, Arthur, 
McNeely, L. R., 
Newton, Robert, 



Reid, Harold W., 
Rudolph, E. G., 
Scudder, H. D., 
Shiffler, C. W., 
Spafford, R. R., 
Taff, p. C, 



Piper, Geo. E., 
Powers, W. L., 



Tucker, Geo. M., 
Wilson, Bruce S., 
WoRRALL, Lloyd. 



AGRONOMIC AFFAIRS. 



Membkr Deckaskd. 
younc, yuncykn. 

Addresses Changed. 

Abbott, John B., R. F. D., Hollows Falls, Vt. 
Allyn, Orr M., Fergus, Mont, 

Beaumont, A. B., Dept. of Agronomy, Mass. Agr. College, Amherst, Mass. 
Bell, N. Eric, Ashville, Ala. 

Briggs, Glen, U. S. Agr. Expt. Sta., Agana, Guam. 

Bushey, A. L., 210 Waldron St., La Fayette, Ind. 

Conrey, G. W., Soils Dept., Ohio State University, Columbus, Ohio. 

Finnell, Howard H., Box 118, Hartshorne, Okla. 

Foord, Jas. a., 3 The Circle, Ithaca, N. Y. 

Longman, O. S., Deloraine, Manitoba, Can. 

Lyness, W. E., Cheyenne Field Station, Archer, Wyo. 

Miles, F. C, Union Grove, Wis. 

Moomaw, Leroy, Havre, Mont. 

Morison, a. T., Connersville, Ind. 

Osenburg, Albert, Scottsbluff Expt. Farm, Mitchell, Nebr. 
Palmer, H. Wayne, 175 12th Ave., Columbus, Ohio. 
Welch, J. S., L. D. S. Maori Agr. College, Hastings, N. Z. 
Zerban, F. W., German Kali Works, Propaganda Dept., 42 Broadway, New 
York, N. Y. 

NOTES AND NEWS. 

John L. Bayles, assistant in agronomy at the Garden City (Kans.) 
substation, has resigned to accept a position in the agricultural de- 
partment of the St. Louis and San Francisco Railroad Company. 

M. A. Brannon has resigned as president of the University of 
Idaho, 

Glen Briggs of the Oklahoma college has been appointed agronomist 
of the United States experiment station on the Island of Guam. 

I. D. Cardiff has resigned as director of the Washington station. 
His successor has not yet been named. 

W. H. Johnson, assistant in the soils section at the Iowa station, is 
pursuing graduate studies at the University of Wisconsin. His work 
at the Iowa station is being conducted by Knute Espe. 

Frank C. Miles, assistant in fiber crops in the U. S. Department of 
Agriculture, has resigned to engage in commercial work in fiber pro- 
duction at Union Grove, Wis. 

A. T. Morison, assistant in crop production at the University of 
Illinois, since April 15 has been county agent in Fayette County, 
Indiana. 



256 



AGRONOMIC AFFAIRS. 



J. A. Ratcliff has resigned as assistant professor of experimental 
agronomy in the University of Nebraska, effective April i, to engage 
in farming in Oklahoma. 

The Secretary of Agriculture has appointed a Committee on Seed 
Stocks, of which R. A. Oakley is chairman. This committee hopes 
to act as a clearing house for seed supplies, particularly State short- 
ages and surpluses. Agronomists who knoAv of surplus stocks of 
seed which are not needed locally but which are likely to be useful else- 
where will confer a favor on the Committee by sending definite in- 
formation to Mr. Oakley regarding them. This information will then 
be transmitted to those who report shortages. The Committee does 
not intend to interfere with State activities looking toward the equit- 
able distribution of seed supplies, but hopes to aid the State organiza- 
tions in locating stocks and in making these available where they are 
most needed. 



JOURNAL 

OF THE 

American Society of Agronomy 



Vol. 9. September, 1917. No. 6. 



ERRATA. 

Wherever the formula (NH,OH),,SO, appears in Tables i and 2, 
read (NHJ2SO4. This error was made by the editor in preparing 
the paper for printing, in attempting to substitute a chemical formula 
for the author's abbreviation " Am. sul." in the tables. The editor 
unfortunately has made little use of chemical formulas in recent 
years and at the time this paper was prepared for publication he was 
on a long field trip on which he did not have access to chemical text- 
books. The error was detected by the author in reading proof, but 
his correction was not forwarded to the editor in the field and conse- 
quently the change was not made before the pages were printed. 
The editor sincerely regrets the error and offers his apologies to the 
author and to the reader. 

c. w. w. 

Sept. 29, 1917. 



A new wheat recently distributed in a limited way by the Kansas 
station has been giving such consistently good results in comparison 
with the standard varieties generally grown throughout the State that 
a brief statement of its yields and milling value, its history, and the 

^ Contribution from the Kansas Agricultural Experiment Station. Received 
for publication February 24, 1917. 



257 



256 



AGRONOMIC AFFAIRS. 



J. A. Ratcliff has resigned as assistant professor of experimental 
agronomy in the University of Nebraska, effective April i, to engage 
in farming in Oklahoma. 

The Secretary of Agriculture has appointed a Committee on Seed 
Stocks, of which R. A. Oakley is chairman. This committee hopes 
to act as a clearing house for seed supplies, particularly State short- 
ages and surpluses. Agronomists who know of surplus stocks of 
seed which are not needed locally but which are likely to be useful else- 
where will confer a favor on the Committee by sending definite in- 
formation to Mr. Oakley regarding them. This information will then 



JOURNAL 

OF THE 

American Society of Agronomy 



Vol. 9. September, 1917. No. 6. 



A NEW WHEAT FOR KANSAS. 

W. M. Jardine. 

Introduction. 

Attempts to improve farm crops by selection and breeding so often 
prove disappointing that the discovery or production of a really su- 
perior strain is always welcomed. Frequently a strain which prom- 
ises well the first few years fails to sustain its record in later years, 
and conversely, many which show little promise in the beginning give 
good results later. In other words, while it is comparatively easy to 
.isolate strains which are superior to the best varieties in certain points, 
it is very seldom that one is found which proves superior in all essen- 
tial respects and which, therefore, under a wide variety of conditions, 
produces better yields or better quality of grain, or both. 

Moreover, it is not easy to prove the superiority of a strain which 
fluctuates widely in yield or quality from year to year with respect to 
other varieties. If it yields better than other varieties in some 
seasons, it is not safe to say that it is the best variety, even though the 
average for several years is significantly higher, since the compara- 
tive value will depend on the frequency of those seasons to which it 
is especially adapted and this cannot be determined without testing 
for a longer period of time than is usually done. 

A new wheat recently distributed in a limited way by the Kansas 
station has been giving such consistently good results in comparison 
with the standard varieties generally grown throughout the State that 
a brief statement of its yields and milling value, its history, and the 

^ Contribution from the Kansas Agricultural Experiment Station. Received 
for publication February 24, 1917. 

257 



258 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

methods by which it was obtained may prove of interest and value to 
Others engaged in similar lines of work. 

Origin and History. 

The new wheat, which has been named Kanred/ is the product of 
a single head selected in 1906 from Crimean (No. 1435 of the Office 
of Cereal Investigations, United States Department of Agriculture) 
by the Department of Botany of the Kansas State Agricultural Col- 
lege. In the fall of 1906, 554 head selections were sown and the fol- 
lowing season 451 were harvested. These and 79 additional selections 
were sown in the fall of 1907, each strain in a single row with alter- 
nate check rows of Kharkof. Five hundred and thirty-three selec- 
tions were harvested in 1908; 122 of these were selected for increase. 
Ten rows of each were sown in the fall of 1908, each row alternating 
with Kharkof as in the preceding season. Eighty-nine of these selec- 
tions were harvested, sown in nursery rows as before, and turned 
over to the agronomy department with about 100 other strains in June, 
1910, for further trial. 

From 191 1 to 1916, inclusive, several of the most promising selec- 
tions, including Kanred, were grown in field plots on the agronomy 
farm. Single plots were grown in 1911 and 1912. Beginning with 

2 Kanred wheat is a product of the Kansas Agricultural Experiment Station 
and can not be credited to any single individual. It was selected by Professor 
H. F. Roberts of the botany department in June, 1906, with over 500 others and 
was grown by him in nursery rows in 1907, 1908, and 1909. It was transferred 
to the agronomy department in 1910 with nearly 200 other strains, 89 of which 
were thought to be of superior merit, for further trial and has been grown by 
that department and its farmer cooperators until the present time. 

The author in the capacity of agronomist took charge of the cereal crop im- 
provement work in June, 1910, when the present work except as noted above 
was started. First as agronomist and later as director of the station it has 
been his special care to keep the various phases of the work correlated and 
allow none of them to lapse because of neglect or changes in personnel. 

Credit is due to Professor L. E, Call, who, as agronomist, has supported the 
work in a most commendable way; to Prof. S. C. Salmon, who has had charge 
of the cereal improvement work since 1913 ; to Mr. H. H. Laude, who had 
charge of the details of the cereal improvement work in 191 1 and 1912; and 
to Mr. R. K. Bonnett and Mr. R. P. Bledsoe in the same capacity in 1913 and 
in 1914 and 1915 respectively. 

Special mention is due Mr. C. C. Cunningham and Mr. Bruce Wilson for 
efficient and careful work in conducting the cooperative tests with farmers; 
and to Professor L. A. Fitz and Miss Leila Dunton for conducting the milling 
and baking tests and chemical analysis of the wheat and flour. 

Credit for assistance in the preparation of this manuscript and assembling 
the data used is due to Prof. S, C. Salmon. 



jakdine: a ni-:w wheat for Kansas. 



259 



1913, three fortieth-acre plots of each variety have been grown, the 
plots being distributed over the area used for the test to reduce the 
experimental error. Since 1914 the Kanred has been grown on the 
substation at Hays, Kans., and in cooperative tests with farmers 
throughout the hard winter wheat belt. It has been grown at the 
substations at Colby and Garden City, Kans., since 1915- ^Tilling 
and baking tests and c-hemical analyses of the most promising strains 
at Manhattan have been made each year since 191 2. 

Characteristics. 

The wheat in question is a hard winter variety, characterized by 
the presence of awns, whitish, glabrous glumes, and reddish grain of 
the well-known Crimean or Turkey type. In habit of growth and 
general appearance the plant and grain cannot be distinguished from 
Turkey and Kharkof unless it be in minute botanical differences which 
have not been determined. It usually heads and ripens somewhat 
earlier than Turkey and Kharkof, but this difference is not sufficiently 
constant for identification. 

Yield. 

The early records secured by the Department of Botany indicated 
the superiority of Kanred as compared with the varieties usually 
grown. However, it was not until it had been grown in field plots for 
six years and in cooperative tests with farmers in many parts of the 
State that its value was felt to be fully demonstrated. In both sea- 
sons in which it was grown in nursery rows it yielded considerably 
better than the adjoining check rows, although less than other selec- 
tions grown with it. In 1908 it produced 29.5 percent and in 1909 
over 30 percent more grain than the check rows of Kharkof. The 
difference has been somewhat less when grown in field plots. The 
yields and other agronomic data secured in the tests on the agronomy 
farm with Kanred, Turkey, and Kharkof are shown in Table i. 
The Turkey and Kharkof strains are those which have been grown by 
the college for many years and have been widely distributed to 
farmers throughout the State. 

It will be seen that the average yield of Kanred is 4.6 bushels higher 
than that of Turkey and 5.2 bushels higher than that of Kharkof. 
It has exceeded these varieties in yield in every year but one, and in 
that year the difference was less than the experimental error. In 
nearly every season the new variety has headed and ripened earlier 
than the other varieties mentioned. The average difference is from 
one to over two days. The average weight per bushel of Kanred 



260 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

is 0.3 pound less than that of Turkey, and 0.9 pound higher than that 
of Kharkof. The difference is probably within the error of deter- 
mination. 

The probable error of the yield in each case where more than a 
single plot was grown has been computed and is included in the table, 
but great weight should not be given a probable error determined from 
three plots only as is the case here. Considering the large differences 
in favor of Kanred in 1912 and again in 1916, and the very great 
chances against a variety producing better than two others of like 
yielding ability in five years out of six, there can be little doubt as to 
the superiority of Kanred for this period and for the conditions 
under which it was grown. 



Table i, — Yield and other agronomic data recorded on Kanred, Turkey, and 
Kharkof at Manhattan, Kans., in the six years from igii to 1916, inclusive. 

YIELD IN BUSHELS PER ACRE. 



Variety. 


1911. 


1912. 


1913- 


1914. 


1915. 


1916. 


Average. 


Differ- 
ence 
com- 
pared 
with 
Kanred. 


Kanred . . 
Turkey . . 
Kharkof.. 


34-6 
26.1 


19.8 

13-2 

11.9 


37.i±i-i4 35-2± .96 
33-^^±i-59 36.irb1.09 
33-8± .63 36. o± .71 


26.0 ±.04 
23.oi.05 

22.9±.05 


33-6± .28 

22.2±1.39 

24.6ii.78 


3I-I 
26.5 
259 


-4.6 
-5-2 


DATE HEADED. 


Kanred . . May 21 1 May 17 
Turkey . .1 do 1 May 18 
Kharkof..] do | 1 May 19 


May 22 
May 24 
May 25 


May 24 
May 25 
do 


May 25 
May 26 
do 


May 21.8 
May 22.8 
May 23.1 


I.O 

1.3 


DATE RIPE. 


Kanred . . June 7 | June 17 
Turkey . . June 8 1 June 18 
Kharkof. . do | do 


June 15 
June 17 
do 


June 30 
July 2 
July 3 


June 27 
June 29 
June 30 


June 19 
June 20.8 
June 21.2 


1.8 
2.2 


WEIGHT PER BUSHEL, POUNDS. 


Kanred . . 
Turkey . . 
Kharkof.. 


60.5 62.5 1 57.0 
62.0 60.5 57.9 
61.8 1 57.3 1 59.3 


59.0 
59.6 
58.4 


57-0 
55-0 
54-0 


59-3 
60.2 
59-0 


59-2 
59-5 
S8.3 


•3 

- -9 



Very similar results were obtained in tests at the substations. The 
yields for Kanred at the substations, compared with Turkey or Khar- 
kof, or both, are shown in Table 2. 

No yield record was obtained at Hays in 191 5, as extremely wet 
weather caused the grain to lodge so badly that accurate data could 
not be gotten. In each case the best strains of Turkey and Kharkof 



jardine: a nrw wheat for Kansas. 



261 



grown at each station are included for comparison. In every test 
and in every year but one, Kanrcd outyielded Turkey and Kharkof. 
The average difference at Hays is 3 bushels, at Colby 2.8 bushels, and 
at Garden City 1.7 bushels. At Garden City in 1915 Kharkof and 
Kanred produced practically the same yield. 



Table 2. — Annual and average yields in bushels per acre of Kanred, Turkey, 
and Kharkof wheats at three substations in Kansas, 1914 to igi6. 



Station and variety. 


Yield in bushels per acre. 


1914. 




1916. 


Average. 


Hays, Kansas: 












25.6 




36.4 


31-0 




233 




32.7 


28.0 


Garden City, Kansas: 










Kanred 




"15-4 


17.2 


16.3 


Turkey 




13.8 


15-3 


14.6 


Kharkof 




15.5 


13-3 


14.4 


Colby, Kansas: 










Kanred 




34-3 


42.6 


38.5 


Turkey 




33.8 


28.8 


31-3 



o Average of three tests. 



Kanred wsls included in ten tests with farmers in the hard winter 
wheat belt in 1914, in 25 tests in 191 5, and in 21 tests in 1916. In 
most cases Turkey and Kharkof from the college were included and 
also the variety used for seeding the general fields of the farmer con- 
ducting the test. In all cases this local variety was a hard winter 
wheat of the Crimean type and in many cases was the product of 
seed originally obtained from the college. These tests were conducted 
by seeding a plot one or two drill widths wide and of convenient 
length. The yields were determined by the hoop method to be de- 
scribed later. On account of the considerable space required the 
tabulated data showing the results in detail are omitted here. 

In 1914 Kanred produced a higher yield than the local variety in 
nine tests of the ten that were conducted, and in the tenth, the dif- 
ference was less than a quarter bushel. It also exceeded Turkey 
and Kharkof in every test but one and in this the difference was 
small. The average for all tests in 1914 show a gain of 3.4 bushels 
as compared with the local variety, of 2.5 bushels compared with 
Turkey, and of 3.2 bushels as compared with Kharkof. ^ 

In 191 5 Kanred produced a higher yield than the local variety in 19 
of the 23 tests which included both; higher than Turkey in 18 of 23 
tests; and higher than Kharkof in 19 out of 24 tests. The average 
of those tests which included all four varieties shows a gain of 3.5 



262 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



bushels compared with the local variety, 2.0 bushels compared with 
Turkey, and 4.5 bushels compared with Kharkof. The results for 
191 5 are somewhat more erratic than in 19 14, due perhaps to the ex- 
cessively wet season and considerable damage from disease and 
lodging. 

There was but one test in 1916 in which a distinctly larger yield 
was secured from the local variety than from Kanred. In two tests 
practically the same yields were secured from each, while in 18 of the 
21 tests conducted Kanred produced decidedly the most grain. 
Kanred gave higher yields than Turkey in all tests but two and in one 
of these the difference was well within the limits of experimental 
error. It produced higher yields than Kharkof in every test but one. 
The average gain was 6.2 bushels compared with the local variety, 5.9 
bushels compared with Turkey, and 6.5 bushels compared with 
Kharkof. The superiority of Kanred was especially apparent in the 
north central and northeastern counties, where considerable winter 
injury to other varieties occurred. 

To summarize, Kanred has given an average gain compared with 
Turkey of 4.6 bushels at Manhattan, 3.4 bushels in all tests at the 
substations, and 3.6 bushels in cooperative tests with farmers, or an 
average difference for all tests of 3.7 bushels. It has exceeded Khar- 
kof by 5.2 bushels at Manhattan, 5.1 bushels in cooperative tests with 
farmers, and an average of 4.7 bushels for all tests. In the 54 co- 
operative tests with farmers which included the local variety, it gave 
an average increase of 4.4 bushels over the local strain. In all tests 
conducted it has exceeded Turkey in yield 59 times out of a possible 
66, Kharkof 51 times out of a possible 58, and the local variety grown 
in cooperative tests 49 times out of 54. 

Earliness and Cold Resistance. 

Kanred has headed and ripened on the average fully a day earlier 
than Turkey and more than a day earlier than Kharkof at Manhattan. 
It was observed in some of the cooperative tests that this strain was 
among the earliest to mature. In other cases it ripened at the same 
time. 

There are strong indications that Kanred is able to survive severe 
winters better than other varieties. In 1912, which was the only 
season in which winterkilling occurred at Manhattan, notes taken in 
the spring show that 90 percent of Kanred plants survived as com- 
pared with 80 percent of Turkey and 77 percent of Kharkof. In the 
spring of 19 16, Kanred was noticeably more vigorous in appearance 



jardinr: a new wiikat for Kansas. 263 

than other varieties. No severe winter injury occurred at Manhattan, 
but in the north central and northeastern parts of the State consider- 
able injury occurred as a result of a heavy covering of ice during a 
part of the winter. In cooperative tests in this area, Kanred survived 
very much better than other varieties. This ice sheet was also present 
at Manhattan and may have been responsible for the large difiference 
in yield between Kanred and other varieties, even though no imme- 
diate effect on the plants was noticeable. 

It seems probable from experiments conducted at this station that 
winter injury may be due to several factors and that a variety re- 
sistant to one factor is not necessarily resistant to another. Hence, 
it would not be safe to infer that Kanred will prove hardier than other 
varieties with different conditions. 

Milling and Baking Tests. 

Mining and baking tests of Kanred have been made since 1912. 
Table 3 presents the most important data in comparison with similar 
figures for Kharkof and Turkey in 1912, 1913, 1914, and 1915. The 
results for 1916 were not available when this paper was written. 



Table 3. — Data recorded on milling tests of wheat and baking tests of flour from 
Kanred, Kharkof, and Turkey wheats at Manhattan in igi2, 1913, igi4,and 191 5. 





Wh 


eat. 




Flour. 


c 






V 










a 


a 

V 


Year and variety. 


Protein 
con- 
tent. 


Moist- 
ure con- 
tent. 


Yield c 
flour. 


Protein 
con- 
tent. 


Moist- 
ure con- 
tent. 


Absorpti 


Maximu 
expansic 


Oven ris 


Loaf 
volume 


Color. 


Textur 


Weigh 


Wet glut 


Dry glut' 


I912 : 

Kanred 


% 


% 


% 


% 


% 


% 


CC. 


cm. 


CC. 


% 


% 


gm. 


% 


% 


17.40 


9.89 


67.08 


17.02 


11-34 


58.3 


2400 


6.0 


1940 


94 


93 


518 


57.66 


16.43 


Kharkof 


17-38 


10.84 


65.40 


16.90 


11.67 


60.0 


2350 


6.0 


1900 


96 


97 


523 


57-93 


16.60 


1913: 
































17.02 


8.32 


65.68 


15-15 


12.77 


61.7 


2100 


5-1 


1910 


93 


96 


531 


49-38 


14.89 


Kharkof 


11.38 


10.84 


65.04 


9.34 


13.92 


62.7 


2250 


3-3 


1760 


92 


95 


532 


29.02 


9.12 


Turkey 

1914: 


13.96 


8.38 


66.88 


12.43 


12.52 


65.0 


2100 


4-3 


1880 


91 


94 


534 


37-17 


11.93 


15.15 


12.19 


64.68 


14.38 


13.75 


58.5 


1900 


4-7 


i860 


91 


90 


524 


44-73 


14.52 


Kharkof 


15-75 


12.07 


63.80 


14.57 


13.39 


60.3 


1800 


4-2 


1820 


89 


89 


538 


46.09 


14.31 


Turkey 


14-37 


11.36 


65.16 


13.04 


13.43 


59-4 


1600 


4.0 


1800 


87 


89 


525 


38.13 


12. II 


1915: 
































20.78 


13.02 


59.10 


18.22 


13.94 


60.0 


2200 


6-5 


2040 


93 


94 


513 


58.41 


17.58 


Kharkof 


18.8s 


12.78 


58.80 


16.48 


13.73 


60.0 


2350 


6.1 


2030 


93 


91 


521 


50.56 


15.90 


Turkey 


20.00 


12.34 


61.30 


17.58 


13.24 


60.9 


2350 


6.6 


2095 


90 


94 


500 


51-58 


16.17 


Ave. 1912-15: 
































17.59 


10.85 


64.13 


16.19 


12.95 


59.6 


2150 


5-6 


1937 


92.7 


93-2 


521 


52.54 


15-85 


Kharkof 


15.84 


11.63 


63.26 


14.32 


13.18 


60.7 


2162 


4.9 


1877 


92.5 


93-0 


529 


45-90 


13-98 


Ave. 1913-15: 






























Kanred 


17.65 


II. 17 


63.19 


15.92 


13-49 


60.1 


2067 


5-4 


1937 


92.6 


93-3 


523 


50.84 


13-66 


Kharkof 


15.33 


11.89 


62.54 


13.46 


13-68 


61.0 


2133 


4.5 


1870 


91.3 


91.6 


530 


41.89 


13-I1 


Turkey 


16. II 


10.69 


64-45 


14-35 


13-06 


61.7 


2017 


4-9 


1925 


89.3 


92.6 


520 


42.26 


13.40 



264 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Table 3 shows a distinctly higher protein content both in the wheat 
and in the flour of Kanred than of Turkey and Kharkof ; a higher 
percentage of flour than Kharkof, but somewhat less than Turkey; a 
loaf expansion practically equal to Kharkof and slightly greater than 
Turkey; color of loaf equal to or better than either of the standard 
varieties ; and texture of loaf equal or superior to either of the other 
varieties. The amount of dry and wet gluten is also distinctly in 
favor of the new variety. Stated in other words, there is no im- 
portant point in which Kanred is shown to be inferior to the standards 
used and in some points, notably in protein and gluten content, it 
appears to stand distinctly higher. There would seem to be no ques- 
tion regarding its milling and baking value. 

Crop Improvement Methods. 

It is difficult if not impossible to emphasize any one phase of the 
methods used that is responsible for Kanred. It is a product of the 
pure-line method of selection first used at this station by the Depart- 
ment of Botany. This method permits testing a much larger number 
of varieties and strains than was possible with the older method of 
continuous selection, and this without doubt has contributed largely 
to whatever success has been attained. Anotherpoint of considerable 
importance is the reduction of experimental error to the greatest ex- 
tent possible by replication of plots and use of check plots and by 
exercising all possible care in the use and management of the land de- 
voted to experimental work, in laying out plots, and in seeding, har- 
vesting, and thrashing. 

Land used for field plot work is handled in a definite rotation of 
(i) wheat, (2) corn, and (3) oats. Cowpeas are sown after the 
wheat as a catch crop and plowed under the following fall or spring 
for corn. No manure or commercial fertilizer has been used. The 
area used for the wheat plots is cropped uniformly to corn and oats 
the preceding two years, all experimental plots of corn and oats be- 
ing on a different area. The nursery wheat plantings are alternated 
with a crop of oats and Canada field peas cut for hay. In all cases 
the ground is plowed early and thoroughly prepared before seeding. 
Seeding is delayed until danger from Hessian fly is practically past, 
but early enough to secure a good growth before winter. 

In recent years nursery plantings have been replicated four times 
and field plots three times. Previous to this frequent checks were 
used. The outer drill rows of field plots have been removed before 
harvest to eliminate the alley effect. No provision has been made in 
nursery plantings to prevent the possible effect of adjacent rows, but 



JARDINi:: A NEW WHEAT EOR KANSAS. 



265 



in recent years this effect has been reduced to a minimum by grouping 
all varieties and strains and growing those of like habits of growth 
together. In case of severe winter injury or other damage which is 
likely to affect adjacent rows, the possible effect is considered in in- 
terpreting the results. The heads of grain of all nursery rows are 
wrapped in paper to prevent loss and mixture, and thrashed in a 
specially constructed machine for the same purpose. The grain is 
reweighed to prevent possible mistakes. 

Farmers' Cooperative Tests. 

A factor probably more important than any yet mentioned is the 
practice first developed at this station of thoroughly testing supposedly 
better strains with the farmers themselves before any wide distribu- 
tion of seed is made. To the writer this seems of first importance, 
as the most accurate tests conducted at one place in a State with the 
diversity of soil and climate possessed by Kansas may mean nothing 
for other localities. Tests at the experiment station indicate some- 
thing of the value of varieties for other localities, but they cannot 
prove their worth and it is a mistake to assume that they do. 

This idea is not a new one in plant-breeding work, but it has never 
been emphasized as much as others of probably less importance. 
For example, in the German method of breeding, the principal method 
in vogue until less than 15 years ago, the main emphasis was placed 
on the selection of the best individuals each year for the propagation 
of an elite strain the following year. It was assumed that the im- 
proved varieties produced in this way would yield better than the un- 
selected varieties, but very little effort was made to demonstrate the 
truth of this assumption. In the pure-line method which has prac- 
tically supplanted the German method in this country, more attention 
has been given to determining the worth of the different selected 
strains, but in this method the necessity of testing the improved strains 
for all conditions to which they are supposed to be adapted has not 
been emphasized. 

In Kansas improved varieties are sometimes sent into localities in 
which no formal test has been made, but only with the distinct under- 
standing that they are sent for trial only and not with the unqualified 
recommendation of the experiment station. 

Methods of Conducting Cooperative Tests. 

The method used in making tests in cooperation with farmers is 
important and may be of interest. In the first place, only those 
farmers who are willing and able to give some extra time to conduct- 



266 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

ing the test are chosen. To these are sent about 15 pounds of each 
variety to be tested. Each test includes from four to ten varieties, 
one of them being the variety used by the farmer for general seeding. 
The remaining varieties include only those known in a general way 
to be adapted to the region. At seeding time a single drill-width or 
two drill-widths of each variety are sown, the length of the plot con- 
forming either to the size of the field, the area of uniform soil avail- 
able, or to the quantity of seed. The varieties are sown side by side 
with alleys 12 to 18 inches wide between plots. At harvest time a 
representative of the college visits the cooperator, takes such data as 
can be obtained at that time, and harvests a representative area of 
the plot for yield. The tests are inspected at other times of the year 
also as far as practicable. The yield is determined by harvesting 
from each plot ten areas inclosed by a hoop approximately 42 inches 
in diameter or a total area of 0.05 acre. The grain and straw when 
harvested are placed in a large jute sack with a tag inside and one out- 
side and are then shipped by express to the college, where the grain 
is dried and later thrashed in the nursery thrasher. 

Considerable care and judgment in choosing the areas and in har- 
vesting the grain inclosed by the hoops is necessary to secure accurate 
results. Tests conducted at various times to check the accuracy of 
this method show that with reasonable care in choosing the hoop areas 
and harvesting them, the results are probably as dependable as those 
obtained by harvesting the entire plot. For example, a 0.4-acre field 
of alfalfa in 191 3 estimated by this method to yield 1.5 tons actually 
produced 1.45 tons, and a 045-acre field of oats estimated at 29.4 
bushels produced 28.3 bushels per acre. Hoop areas were taken from 
eight tenth-acre plots on the agronomy farm in 19 14 and the plots 
then harvested in the usual way. The extreme variation in yield of 
grain secured by the two methods was 5.1 percent and the average 
0.93 percent of the yield of the plots. In 75 'percent of the plots the 
hoop method gave the highest yield, probably because of less loss in 
harvesting. 

Kansas Agr. Expt. Station, 
Manhattan, Kansas. 



IIALSTKD & OWEN: CROWTII OF MATZK SKF.DM N( IS. 



267 



INFLUENCE OF POSITION OF GRAIN ON THE COB ON THE 
GROWTH OF MAIZE SEEDLINGS.^ 

Byron D. Halsted and P^arle J. Owen. 

Introduction. 

Five sample ears from each of 20 representative varieties and 
crosses of corn were used in this experiment. The grains from each 
ear were divided into 10 equal lots, each representing a zone of the 
ear, ranging from the butt to the tip. The average weight of the 
grains in each zone of 2 of the 5 ears was obtained and then 25 kernels 
from each lot were planted an inch deep in a greenhouse bed. The 
grains for planting were taken at random, except that in the butt and 
tip zones the smallest perfect kernels were used. The plants were 
harvested after 17 days, a period sufficient to produce plants of con- 
siderable size. 

The emergence of the tips of the seedlings was recorded daily, and 
from this the averages were obtained. At harvest time the viability 
was recorded and the weight and length of the seedlings taken, from 
which the vigor and variability were afterward deduced. By averag- 
ing the five units (ear belts) in each set a series of tables has been 
constructed that shows the relationship of the position upon the cob 
to the particular characters under consideration, as follows: (i) 
Weight of grain; (2) specific gravity of grain, (3) emergence of 
seedlings; (4) viabihty of seeds; (5) weight of seedlings; (6) length 
of seedlings, and (7) variability in length. 

Weight of Grain. 

The names of the varieties used in the experiment and the weights 
of the grains in grams by zones are shown in Table i. 

The grains range in weight from those of the Hickory King, which 
are unusually large, to those of the Country Gentleman and the Golden 
Queen (pop), the immature kernels of the last being very small. The 
flint varieties generally have the heaviest grains, followed by the 
dent sorts, but the Black Mexican, Stowell Evergreen, and Golden 
Bantam, three standard sweet kinds, are near the average of the 

1 Contribution from New Jersey Agricultural Experiment Station. Received 
for publication January 5, 1917. 



268 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

list of 20 here given. The cross between Hickory King and Golden 
Queen has a rank that is practically the mean of that of the two 
parents. Nearly the same is true of the cross of Golden Queen upon 
Champion White Pearl, and also of Golden Queen upon Brazilian 
Flour. 



Table i. — Average weight of grains {in grams) of 20 varieties of corn, in each 
of 10 Bones from the butt to the tip of the ear. 



Variety. 


Zone. 


Aver- 


Butt 
(I). 


2. 


3. 


4- 


5- 


6.] 


7- 


8. 


9- 


Tip 
(10). 




0.43 


0.44 


0.43 


0.42 


0.42 


0.40 


0.39 


0.37 


0.35 


0.31 


0.396 




.52 


.55 


.56 


•53 


•53 


•51 


.48 


.46 


.46 


.42 


•511 


Golden Queen (mature) 


.16 


.17 


.16 


.16 


.15 


.12 


.13 


.13 


.12 


.10 


.140 


Golden Queen (imma- 


























.06 


.06 


.07 


.07 


.07 


.07 


.06 


.05 


.04 


.03 


•058 


Champion White Pearl. 


.40 


.39 


.38 


.37 


.37 


•35 


•35 


•34 


•33 


.29 


•357 




.29 


.29 


.28 


.28 


•27 


.26 


.24 


.22 


.20 


• 17 


.250 




.32 


.32 


•32 


.31 


.31 


.29 


.28 


.27 


• 27 


.24 


.293 


T^Ann^^ r^rintrfxT" \A/Viiffi 


AO 


•?8 


-^8 


•0 / 


1 c 
'jj 


•«54 - 


•00 


"32 


"3 T 
•0 ••■ 


.28 


.346 




.40 


.38 


.36 


.36 


•36 


.36 


•36 


.34 


•33 


.28 


•353 




•34 


.35 


.35 


•34 


•32 


•32 


•32 


•31 


•31 


.29 


.325 




•30 


.31 


.31 


•31 


•31 


•30 


.29 


.28 


.27 


.24 


• 294 


Stowell Evergreen 


.30 


.28 


.27 


.26 


.26 


•25 


•25 


.24 


.23 


.21 


• 255 


Black Mexican 


.28 


•30 


.29 


.29 


.29 


.28 


.27 


.26 


.23 


.20 


.269 


Country Gentleman. . . 


.18 


•IS 


.15 


.14 


.14 


.13 


•13 


.14 


.12 


.11 


• 139 




.26 


.27 


.27 


•27 


.26 


.26 


.24 


•23 


.21 


.18 


•245 




.25 


.25 


.24 


.23 


•23 


.22 


.20 


.20 


.18 


.17 


.217 


Golden Queen X 


























.26 


.28 


.29 


.28 


.28 


•27 


•27 


•25 


.23 


.18 


.259 


Squaw X Country- 


























•25 


.25 


.25 


.24 


.24 


•23 


•23 


.22 


.22 


.18 


.231 


Golden Queen X 
























Champion White Pearl 


•27 


.27 


.26 


.27 


.26 


•25 


.24 


.22 


.20 


.18 


.242 


Golden Queen X 
























Brazilian Flour 


.24 


.24 


.24 


.23 


.23 


.22 


.21 


.21 


.20 


.18 


.220 


Totals 


5-91 


5-93 


5-86 


5.75 


5^65 


5^45 


5^30 


5.08 


4.81 


4.24 




Averages 


.296 


.297 


.294 


.288 


.283 


.272 


.265 


•254 


.241 


.212 



The tip zone of grains in all the 20 sets is the lightest, but in three 
kinds, Country Gentleman, Crosby Early, and Stowell Evergreen, 
the difference is small. In these the ears were unusually well filled over 
the tip. In the Country Gentleman, for example, the grains at the 
tip are noticeably broad but short and not of the " shoe peg " type 
characteristic of the grains in the middle of the ear. The same is 
true of the butt grains, which weigh more than those adjoining them 
in the second zone. This same peculiarity of broad, flat grains is met 
with in the Champion White Pearl, Boone County White, and Wing 
lOO-Day. Quite generally, many of the ovules near the butt fail to 



IIALSTED & OWEN: CKOWTH OK MAIZE SEEDLINGS. 269 

grow, possibly because of inherent weakness, failure to become fer- 
tilized, or because they are destroyed by the extra pressure of tlic 
closely fitting husks. Therefore, the grains that are formed arc 
flat-topped, misshapen, and arranged without apparent order upon 
the cob. Had all the ovules at the butt developed, it is self evident 
that, conditions remaining the same, the grains would have been 
much smaller than they were found to be. This tendency of grains 
to expand when lateral pressure is removed has been shown by the 
larger size and nearly spherical shape of grains produced on ears 
that for this particular purpose had received very limited amounts 
of pollen, and, as a consequence, produced only a few scattered 
grains. 

It is seen that the zone next above the butt leads in weight, with 
the butt zone a close second, and that the decrease is quite uniform 
from the third zone to the tip. A long ear that tapers both ways 
from the second zone would visually represent the results that have 
been obtained. 

Specific Gravity. 

The range in specific gravity is from 1.35 in Golden Queen to 
1. 1 6 in Brazilian Flour. In the first kind the grains are small, and 
the endosperm is chiefly corneous, while in the last the grains are 
large, with a chalky or floury endosperm. The sweet corns have a 
high specific gravity. Golden Bantam and Stowell following close to 
the Golden Queen, with Country Gentleman, Black Mexican, and 
Crosby not far behind. This high range of specific gravities in the 
sweet grains is doubtless due to the low starch and high sugar con- 
tent and the consequent drying down of the endosperm into the horny 
texture, which is quite comparable with that of the pop and flint 
corns. The crosses between Golden Queen and Brazilian Flour, 
the highest and the lowest in the list for specific gravity, give nearly 
the mean of the two extremes, namely, 1.26. The cross of Golden 
Queen upon Champion White Pearl gives 1.26, which is close to the 
specific gravity of the mother parent, and in like manner the Golden 
Queen upon Hickory King cross shows a specific gravity quite close 
to that of the Hickory King. 

There is a remarkable difference between the specific gravity of 
mature and of immature grains of the Golden Queen. The immature 
grains were the lightest with the exception of the Brazilian Flour 
previously mentioned. In other words, the specific gravity of the 
Golden Queen grains fell from the first place to the last but one 
because of immaturity. This test illustrates the importance of select- 



270 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

ing for the present study only ears that were well filled and showed 
no sign of immaturity ; and with the single exception purposely made, 
this was done with extreme care. 

The sixth zone from the butt is highest in specific gravity, with 
the fifth and fourth zones next in order, followed closely by the third 
and the seventh zones. The three lowest specific gravities are at the 
tip end, and the fourth lowest at the butt. In a general way the 
specific gravity decreases in both directions somewhat regularly from 
the middle of the ear. By making five groups from base to tip, the 
sums of the ranking figures are as follows: 9, 15, 19, 9, 3. This 
shows that the decrease is much more rapid in the upper than in the 
lower half of the ear. When three groups only are made, namely, 
the basal three, the middle three and the upper four zones, the 
averages of the specific gravities are: 1.25, 1.26, and 1.24. 

The ranking of the specific gravities does not coincide with that 
for weights of grains, as the following parallel display shows :2 





I 


2 


3 


4 


5 


6 


7 


8 


9 


10 


Rank of weight of grains 


9 


10 


8 


7 


6 


5 


4 


3 


2 


I 


Rank of specific gravity 


4 


5 


7 


8 


9 


10 


6 


3 


2 


I 


Sums 


13 


15 


15 


15 


15 


15 


10 


6 


4 


2 



Only the three upper zones show the same rank for both weight 
and density. The three lower zones, while the heaviest, have a 
medium specific gravity, and the next three with the greatest specific 
gravity are only medium in weight. 



Emergence. 

As the temperature of the air and soil of the greenhouse was not 
constant during the winter months when the tests were conducted, 
no rational comparison of the emergences is possible. If one dis- 
regards the variation in temperature, it follows that the Wing 1 00- 
Day, Early Teaming, Champion White Pearl, and Iowa Silvermine 
are quick growing, while, on the other hand, Brazilian Flour, Golden 
Queen, Hickory King, and Longfellow are comparatively slow. 

From the grand averages for all the zones it is found that the time 
consumed from the planting to the showing of the tip at the surface 
of the ground is somewhat less than one week (6.44 days). The 
butt zone required the shortest time for emergence, a trifle over 6 

2 The number 10 stands for highest and from it the series descends in regu- 
lar order to i, the lowest. 



HALSTKH & OWEN : GROWTH OK MAIZK Si;i:i )r J N( IS. 2/ I 

days, while the zone next above it needed lo hours longer, and was 
the slowest zone upon the ear. It is further found that there is a 
fairly uniform decrease in time for emergence from the second to 
the tenth zone. When the ranks of emergence are compared with 
those of weight of grain, it is seen that there is a close correspond- 
ence, that is, a high weight is associated with a low rate of emergence. 
This is shown in the following table, which gives the ranking of the 
ten zones for all the lOO ears. 



Zone 


I 


2 


3 


4 


5 


6 


7 


8 


9 


10 


Rank of weight of grains 


9 


10 


8 


7 


6 


5 


4 


3 


2 


I 


Rank of emergence of seedlings .... 


I 


10 


9 


8 


6 


7 


5 


2 


4 


3 


Sums 


10 


20 


17 


IS 


12 


12 


9 


5 


6 


4 



The exception to the general rule is found in the butt zone, where 
the grains rank next to the largest, and their seedlings were the first 
to emerge. This result is accounted for in part by the fact that the 
smallest grains of this zone were planted. Furthermore, the kernels 
were much compressed and misshapen and therefore they exposed 
an absorbing surface relatively, if not actually, greater than that of 
the grains from the other portions of the cob. 

Viability. 

The viability of the grains of a large portion of the samples was 
high, the general average being 91.91 percent. This includes the set 
of immature ears of the Golden Queen, which had a viability of 67.87 
percent, with Brazilian Flour only a little better (76.94 percent) and 
Stowell Evergreen not far ahead. The most viable sets were Iowa 
Silvermine and Early Leaming, both with 98.72 percent viability, fol- 
lowed closely by a cross of Golden Queen upon Champion White 
Pearl (97.92 percent), the viability of the parents being nearly the 
same, namely, 96.96 percent and 96.56 percent, respectively. With 
the exception of the Brazilian Flour (and, of course, the immature 
Golden Queen), the sweet corns as a group were the least viable, a 
result that has always been observed in previous comparative studies 
of viability in corn. 

The results show further that the greatest viability is in the fourth 
zone, with an average of 9444 percent, and the lowest at the butt 
(87.68 percent), followed closely by the tip. The best five zones for 
ability to produce seedlings are contiguous in the middle of the ear, 
leaving two at the butt and three at the tip to compose the poorer half. 



2J2 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



The ears were too well selected to yield any striking differences in 
viability as associated with place upon the cob. 

The following tabulation of the ranking figures shows that the 
weights, specific gravities, and viabilities are associated. 





I 


2 


3 


4 


5 


6 


7 


8 


9 


10 




9 


10 


8 


7 


6 


5 


4 


3 


2 


I 




4 


5 


7 


8 


9 


10 


6 


3 


2 


I 




I 


3 


4 


10 


7 


9 


6 


8 


5 


2 


Sums 


14 


18 


19 


25 


22 


24 


16 


14 


9 


4 



The heavier grains are in the lower half of the ear, but those of 
greater density and viability occupy the middle zones of the cob. 

Vigor. 

Vigor is here recorded as the live weight of the seedlings from 
which that of the grain itself has been deducted. An error is intro- 
duced into this method by the fact that the original grain coats, con- 
taining much of the original food substances and considerable water, 
adhere to the seedHngs. It was found too difificult to remove these 
watery kernels before weighing, and therefore the results are all 
somewhat too high, but the error in all instances is probably a fairly 
uniform one. 

Furthermore, there is a variable time factor that should be men-| 
tioned, namely, that there are several hours' difference between the 
emergence of the seedlings of the smaller grains and those of the 
slower growing plants from the larger grains upon the same ear, 
usually those near the tip and butt. This difference in germination, 
therefore, favors the seedlings from grains borne at the extremities 
of the ears, and makes the recorded differences somewhat less than 
they should be. 

A comparison of the ranking columns of the size of grains with 
those of seedling vigor shows a positive correlation. The data on 
live weight of the seedlings are given in Table 2. 

From Tables i and 2 it is seen, for example, that the Hickory 
King is first in weight of grain and vigor of seedlings, while the im- 
mature Golden Queen is lowest in both these characters. Contrari- 
wise, the mature Golden Queen, among the lightest of all the mature 
sets, is next to the lowest in vigor of the seedlings. There is a great 
lack of vigor in the comparatively large butt grains as shown in the 
display of ranking figures which immediately follows Table 2. 



IIALSTED & OWEN : GROWTH OF MAIZE SEEDLINGS. 2/3 



Table 2. — Vigor of seedlings as indicated by their live weight in grams, as 
related to the position of the kernel on the cob. 













Zone. 














Variety. 
























"a 


























I 


■ 


3 


4 


5 


6 


7 


8 


9 


10 




< « 


p2 




2,266 


2.386 


2.238 


2.428 




2,444 


2,366 


2,224 


2,328 


1,692 


2,277 


14 




2.870 


3.034 


3.042 


3.276 


1 tin 




3.018 


3,090 


2,888 


2,830 


3.050 


20 


Golden Queen 




























1,000 


1,186 


1.176 


1. 148 


I 182 


1,184 


1,038 


1,094 


1.044 


768 


1,082 


2 


Golden Queen 




























351 


410 


447 


433 


4U/ 




354 


303 


256 


225 


356 


I 


Champion White 




























2,722 


3.120 


2.834 


3.004 


2,960 


2,872 


2,698 


2.694 


2,642 


1,900 


2,744 


16 


Reid Yellow Dent.. 


1,202 


1,468 


1,466 


1.496 


1 ,526 


1 ,468 


1,366 


1,272 


1,214 


916 


1,339 


4 


Iowa Silvermine . . . 


2,282 


2,556 


2,630 


2.682 


2,412 


2,452 


2,444 


2.358 


2,286 


1.638 


2,374 


16 


Boone County 


























White 


2,128 


2,214 


2,250 


2.242 


2,296 


2 1 00 


2,178 


2,156 


2,076 


1,702 


2,134 


12 




2,128 


2,310 


2,234 


2.192 


2 180 


2 286 


2,240 


2,172 


2,916 


1.398 


2,105 


II 




2,140 


2,368 


2.314 


2.220 


2,138 


2.138 


1,980 


2,172 


2,030 


1,748 


1,911 


9 


Brazilian Flour. . . . 


708 


1,074 


1,298 


1,180 




I 202 


1.254 


1,250 


1,280 


952 


1. 155 


3 


Stowell Evergreen.. 


2,662 


2,590 


2,742 


2,568 




2,450 


2,456 


2.350 


2,246 


2,010 


2,447 


17 


Black Mexican. . . . 


1.730 


1,988 


2.058 


2,138 


2 088 




1,970 


1,996 


1,766 


1.352 


1,908 


8 


Country Gentleman 


1,792 


2,336 


2.062 


1,990 


1,904 


1,764 


1,738 


1.638 


1,196 


1,300 


1,832 


6 


Golden Bantam. . . . 


2,178 


2,438 


2.426 


2,590 


2,256 


2,428 


2,466 


2,360 


2,198 


1,652 


2,326 


15 




2,520 


2,596 


2.672 


2,610 


2,624 




2,526 


2,428 


2,466 


2,360 


2,198 


lis 


Golden Queen X 


























Hickory King . . . 


1.546 


1,846 


1.918 


1.930 


1.952 


1,910 


1,884 


1.744 


1.786 


1,258 


1.777 


5 


Squaw X Country 


























Gentleman 


1,688 


2,016 


2,020 


2,064 


1,988 


2,000 


1.938 


1,982 


1,866 


1,422 


1,898 


7 


Golden Queen X 


























Champion White 


























Pearl 


2,070 


2,440 


2.428 


2,436 


2,272 


2,408 


2,204 


2,094 


1.938 


1.430 


2,272 


13 


Golden Queen X 


























Brazilian Flour . . 


1,884 


2,130 


2,064 


2,032 


1,928 


2,036 


1,832 


1,814 


1,810 


1.754 


1,928 


10 




1.893 


2,125 


2,116 


2,133 


2,094 


2,064 


2,010 


1,966 


1,896 


1.503 


1,980 




Rank 


2 


9 


8 


10 


7 


6 


5 


4 


3 


I 









I 


2 


3 


4 


5 


6 


7 


8 


9 


10 


Rank of weight of grains 


9 


10 


8 


7 


6 


5 


4 


3 


2 


I 


Rank of vigor of seedlings 


2 


9 


8 


10 


7 


6 


5 


4 


3 


I 


Sums 


II 


19 


16 


17 


13 


II 


9 


7 


5 


2 



The explanation of this may be found in the unfavorable conditions 
for germ development at the butt. So far as these results go it is 
shown that in respect to vigor the best grains are to be found in the 
lower half of the ear, with the exclusion of the kernels at the butt. 
The upper half shows a uniform decrease in vigor from the sixth 
zone to the tip. 

With the single exception of the butt zone, the weight for the grains 
increases from the butt to the tip. 

The specific gravity is highest in the grains of the middle half of 
the ear. 



2/4 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Summary, 

Table 3. — Ranking of averages of the seven points considered. 



Character. 


Zone. 


I 


2 


3 


4 


5 


6 


7 


8 


9 


10 


Weight of grain 


9 


10 


8 


7 


6 


5 


4 


3 


2 


I 


Specific gravity 


4 


5 


7 


8 


9 


ID 


6 


3 


2 


I 


Emergence . , 


I 


10 


9 


8 


6 


7 


5 


2 


4 


3 


Viability 


I 


3 


4 


10 


7 


9 


6 


8 


5 


2 


Vigor 


2 


9 


8 


10 


7- 


6 


5 


4 


3 


I 


Length of seedhng 


2 


3 


9 


10 


8 


7 


6 


5 


4 


I 


Variabihty 


10 


8 


3 


5 


4 


2 


6 


I 


7 


9 



The emergence is least rapid in seedlings from grains borne in the 
middle of the ear; and is correlated with specific gravity, that is, the 
lighter the grain, the more rapid the initial growth. 

Between viability and vigor there is a strong positive correlation, 
that is, the more viable the more vigorous, and the same is naturally 
true of the length of the seedling. 

Variability is evidently correlated with weakness — but to establish 
the degree, measurements of more than 2,500 seedlings are required. 
In other words, viability and vigor are negatively correlated with 
variability. A set of strong plants is more uniform than one of weak 
seedlings. 

The position on the cob has much influence upon the variability, 
which may be due to size as well as to maturity and nourishment of 
the grains. 

A practical application of the results above given would consist of 
germinating a liberal sample, say 20 kernels, from two rows upon op- 
posite sides near the middle of the ear. Select only those ears that 
show practically 100 percent viability, and plant from only the middle 
of the ear, that is, reject all grains of the butt zone and of the four 
zones of the upper portion of the ear. 

This suggestion does not interfere with any rules in corn selection 
now in practice, but simply is a far more rigid application of the 
method of " nubbing," namely, the discarding of a few grains at the 
butt and tip of the ear, now recommended by corn experts and prac- 
ticed by growers to a limited extent. 



IIAILF.V : HANDLING AND STORAGK OF WIIKAT. 



2/5 



THE HANDLING AND STORAGE OF SPRING WHEAT.^ 

C. H. Bailey. 

The greater part of the small-grain crop of commerce grown in the 
Great Plains area and the eastern part of the United States is handled 
in bulk. Certain factors involved in the successful handling and 
storage of grain vary, depending upon whether it is handled in bulk 
or in sacks. In the first place, there is less opportunity for change in 
the moisture content of bulk grain. Slight reductions in moisture 
content may result when relatively damp grain is handled on a hot 
day, while a slight increase may occur when very cold grain is ex- 
posed in a warm, humid atmosphere. In general, however, the mois- 
ture content of spring wheat is determined principally by the climatic 
conditions prevailing between harvesting and thrashing. If this 
period is warm and dry the grain will be well cured when it starts on 
its journey to the consumer; rain on the unthrashed bundles, par- 
ticularly if exposed in the shock, results in damp, tough " wheat that 
will cause difficulties in handHng and storing. 

The fact that wheat is a relatively poor conductor of heat intro- 
duces another variable in handling wheat in bulk as compared with 
handling in sacks. The heat which develops when damp wheat is 
stored does not pass ofif as rapidly from a large bulk as from a 
smaller one, such as exists when sacks are piled in narrow stacks. 
The more rapid the transfer of heat from a fermenting mass to a 
cooler surrounding medium (usually air) the less the likelihood of 
serious damage. This is assuming that the moisture content of the 
grain is sufficiently low to preclude germination. 

Spring wheat is not biologically ripe at the time it is usually har- 
vested. The post-harvesting process of ripening is attended by cer- 
tain peculiar phenomena. If the bundles are in a stack they take on 
a moist condition. This process is commonly called " sweating," and 
is undoubtedly accompanied by biochemical changes resulting from 
enzymic activities within the kernel. If the sweating process occurs 
in normal wheat in the bin, a slight rise in temperature may result. 

1 Read at the Second Inter-State Cereal Conference, University Farm, St. 
Paul, Minn., July ii, 1916. Published, with the approval of the Director, as 
Paper No. 62 of the Journal Series of the Minnesota Agricultural Experiment 
Station. Received for publication April 20, 1917. 



276 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

The baking quality of the flour is improved by these changes in the 
grain. 

There are several factors which determine whether or not grain 
will spoil in storage after it has passed through the sweat. First, and 
most important, is the percentage of moisture in the kernels when 
they are stored. The form in which moisture exists in the kernel is 
of interest in this connection. Organic colloids of the nature of those 
which form the principal constituents of the wheat kernel have the 
property of imbibing considerable quantities of water and forming 
elastic gels. The colloidal gel swells considerably, although the total 
volume of the water plus the dry colloid diminishes. The water-im- 
bibing capacity of the several colloids varies widely. Thus starch has 
only about one-fourth the imbibing capacity of wheat gluten. There 
is no fixed amount which a given dry colloid will imbibe ; thus gels of 
varying viscosity can be produced, depending upon the proportion of 
water present and upon other variables, such as temperature, mineral 
salts, and other substances. The rate of diffusion in a gel varies 
with the viscosity; in dilute gels diffusion takes place as in water, 
while in strong gels the rate is slower.^ It is probable that in dry 
grain the imbibed water is not sufficient to produce a gel, i. e., the 
colloidal material does not have a continuous structure. The possi- 
bilities of diffusion are decidedly reduced under such conditions. 

The exact percentage of moisture below which this discontinuous 
structure exists is not known; it probably varies with the percentage 
of gluten in the grain, since gluten possesses a greater water-imbibing 
capacity than starch. Increasing the moisture content above the 
maximum at which discontinuity exists results in the formation of an 
elastic gel through which diffusion can occur. Further increases in 
moisture content up to saturation (maximum imbibition) produce 
progressively less viscous gels, and correspondingly increase the rate 
of diffusion. Since the rate of respiration in grain doubtless depends 
in part upon the rate of diffusion between the various kernel struc- 
tures, it follows that the less viscous the gelatinous material of which 
the cell contents are composed the more rapid the production of heat 
through respiration. To restate, the production of heat is dependent 
upon the activity of the oxidases of the kernel, the complex phenom- 
ena being known as respiration. The latter is accelerated by an in- 
crease in the rate of diffusion, which in turn is dependent upon the 
existence of a gel, and the viscosity of that gel. For these reasons 
the moisture content of the grain determines to a considerable extent 

2 Plimmer, R. H. A. Practical organic and biochemistry, p. 386. New 
York, 1915. 



baim:y : handling and stokagk of whkat. 



277 



the liability of heating when bulk grain is stored, and also the rate at 
which the respiration and consequent heating will occur. 

To ascertain the percentage of moisture which spring wheat may 
contain without heating in store, the Minnesota Grain Inspection De- 
partment and the State Boards of Grain Appeals, in cooperation with 
the Division of Agricultural Chemistry of the University of Minne- 
sota, obtained permission from one of the large elevator companies 
of the State to make observations in grain stored by them. About 
twenty lots of wheat were experimented with, containing from 12.76 to 
17.45 percent of moisture. No lot represented less than a carload 
(1,200 to 1,400 bushels). These observations were made through a 
period of more than one year, covering two summer seasons and the 
intervening winter. The data are too voluminous to be given in de- 
tail in this paper. It was concluded that hard spring wheat of reason- 
able plumpness, containing less than 14.5 percent of moisture, is not 
likely to heat when stored under normal conditions in a temperate 
climate, while similar wheat containing 15.5 percent or over of mois- 
ture is practically certain to heat. Between these limits the possi- 
bility of heating depends upon other conditions which are discussed 
later in this paper. 

The rate of heating in its relation to moisture content is shown in 
figure 16. Two lots of wheat are here compared with regard to the 
time required to become actively heating. One carload contained 15.5 
percent, the other 16.5 percent of moisture. They were put in inside 
elevator bins on September 11 and 12 respectively. The lot con- 
taining 15.5 percent of moisture kept 333 days without heating suffi- 
ciently to necessitate turning and cooling, while that which contained 
16.5 percent had to be 
run and cooled in 49 
days. Had these been 
stored in the spring 
or summer, the time 
elapsing before heat- 
ing began would have 
been much shorter, 
but the cold fall and 
winter weather which 
intervened resulted in 
the heat being lost into the air or surrounding material so fast as to 
preclude a rapid rise in temperature. 

This leads to a consideration of the relation of air temperature to 
the rate of heating. The lot of wheat containing 16.5 percent of 



Fig. 16. Graph showing relation between moisture 
content and rate of heating. 



2/8 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



moisture, mentioned in the foregoing paragraph, required from Sep- 
tember 12 to October 31 to rise from 70° to 80° F. and require atten- 
tion. The mean air temperature during this interval was 44.3°. 
Another lot of similar wheat, containing the same percentage of mois- 
ture, had been stored July 28, and the temperature of it rose from 70° 
to 80° F. in II days, the mean air temperature of this interval being 
62.1°. These data are graphically shown in figure 17. The differ- 



so 



70 



40 





July 
































































































% 


eg/' 






















? 

il 
















— 

1 
























-i- 
































1 























10 20 30 40 50 DAY'S 

Fig. 17. Graph showing effect of atmospheric temperature on the rate of 

heating. 

ence in the rate of heating was due to the greater rate of heat loss 
into the cold atmosphere in the fall. While this difference might not 
have been so great in a larger mass of grain, it shows the effect of 
seasonal influences. 

The location of the bin in the elevator may have considerable to do 
with the rate of the loss of heat from the grain. This is shown by 
experiments conducted with a car of wheat containing 17.5 percent 
of moisture which was stored in an outside bin in a steel elevator on 
September 10, and kept until June 10 the following year, a total of 
303 days. Its record is compared in figure 18 with that of a car con- 
taining 16.5 percent of moisture which was put in an inside bin at the 
'Same time and had to be run and cooled in 49 days. 

The initial temperature of the grain is also significant. This is 



BAILKV : IIANDLINC AND STORAGK OK WinCAT. 



979 



shown by the records of bins 32a and 158b. l^>oth were filled at 
almost the same time with wheat containinj^^ 16.5 percent of mois- 
ture. The initial temperature of that in bin 158b, as shown in figure 
19, was 74° F,, while that in bin 32a was 70°. It took the jattcr 
over five times as long to reach a temperature of 80°. These data 



































































































































































































































































y 




















































r 




























































-/ 




























































/ 


























\ 
































































\ 
































/ 




























































/ 




































































































































































































































s 



























































































































































Fig. 18. Graph showing relation of the location of the grain in the elevator 
during the winter months to the rate of heating. 

also illustrate the acceleration of respiration with a rise in tempera- 
ture, the curve being logarithmic in form, and the rate very rapid as 
the temperature approaches 80°. 

When uniformly mixed wheat heats as the result of respiration, the 
highest temperatures 
are usually reached 
near the surface. The 
exact location of the 
warmest portion 
varies with the 
weather. When the 
surrounding air is 
cold, as in midwinter, 
it is usually from 15 
to 20 feet below the 
surface, while in mild 
or hot weather it is 
likely to be at a depth of from 5 to 8 feet. The changes in tempera- 
ture at different depths are shown in figure 20, which gives the record 
of a bin of wheat that was at freezing temperatures on April 4, and 




Fig. 19. Graph showing relation of the original tem- 
perature of the grain to the rate of heating. 



28o JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



was heating on July lo. At this time the temperature of the grain 
surrounding bulb 5, at a depth of 8 feet, was 81° F., while bulbs 4, 
3, 2, and I, which were at depths of 18, 28, 38, and 48 feet, respec- 
tively, were in grain at temperatures of 70°, 65°, 62°, and 62°. The 
larger quantity of oxygen available to support aerobic respiration in 
the kernels near the surface no doubt is responsible for the more 
rapid rise in temperature in that portion; the heat produced in the 
surface layers is lost into the surrounding atmosphere, however, and 
consequently the highest temperature is reached at a depth of a few 
feet. 

The material of which the bin is constructed affects the keeping 

qualities of damp 
grain in just the pro- 
portion that it affords 
heat insulation. The 
four materials used in 
bin construction are 
ranked in heat-in- 
sulating value by the 
leading elevator con- 
struction companies 
of Minneapolis as 
follows: (i) Hollow 
tile, (2) wood, (3) 
concrete, and (4) 
steel. In cold weather 
the better the conduc- 
tor in which the grain 
is stored and the more 
exposed the location, 
the less rapidly will 
damp wheat heat. In 
hot weather the re- 
verse is true, since the 
heat of the air will be 
transmitted to the 
grain through a poor 

























































































































































































































































































































































































































Bulb 
4 



BuuB 
J 



Fig. 20. Graph showing temperatures recorded at 
different depths in a bin of grain. 



insulator, and the rate of respiration accelerated through the result- 
ant rise of temperature. 

To recapitulate briefly, the handling of wheat in bulk introduces 
certain difficulties which do not exist to so great an extent when it is 
handled in sacks. Wheat which is not perfectly ripe when harvested 



COLOR CLASSIFICATION OF WHEAT. 



281 



"sweats" either in the shock, stack, or bin. If normally dry, this 
sweating improves the baking qualities of the flour. The maximum 
limits of moisture which hard spring wheat may contain without 
danger of heating in a temperate climate are between 14.5 and 15.5 
percent. Whether it actually heats or not depends upon several 
factors, including the hardness of the kernels because of the relation 
of kernel density to gluten content, the size or dimensions of the bulk, 
temperature of the atmosphere, initial temperature of the grain, loca- 
tion and consequent exposure of the bin, and the material of which 
the bin is constructed. 

Section of Cereal Technology, 
Division of Agricultural Biochemistry, 
University of Minnesota, 
and 

Minnesota Grain Inspection Department Laboratory. 



THE COLOR CLASSIFICATION OF WHEAT.^ 

The color classification of wheat was discussed by the Minnesota 
Section of the American Society of Agronomy during the first months 
of its organization. The subject was first discussed by H. K. Hayes 
of the Section of Plant Breeding, and a committee was appointed to^ 
devise a scheme which could be employed by all who had occasion 
to use such a system of classification. A set of samples was pre- 
pared by A. C. Arny and P. J. Olson which illustrated the various 
divisions provided for in this system. The report of this committee 
was subsequently adopted by the local section, which authorized its 
pubHcation in the Journal of the society. This report is presented 
in the following paragraphs. 

The visual appearance of wheat which is commonly termed color 
is due to the joint effect of two factors : First, the presence or absence 
of a brownish-red or orange-yellow pigment in the bran layer, and 
second, the physical condition of the endosperm cells. The latter 
may be corneous or starchy, depending upon the density of the cell 
contents or the relative amount of space occupied by air cavities or 
vacuoles. The confusion which has arisen in regard to color classi- 
fication is probably due to the use of a single term to describe the 
combined visual effect of these two characters mentioned above. 

1 Prepared by a Committee of the Minnesota Section of the American So- 
ciety of Agronomy, consisting of Messrs. H. K. Hayes, C. H. Bailey, A. C. 
Arny, and P. J. Olson. Received for publication March 17, 1917. 



282 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

That there is such a confusion is probably recognized by every one 
who has any knowledge of the present methods of classifying the 
color of wheat. 

The present state of ideas is illustrated by the test conducted by 
H. K. Hayes. Six samples of wheat were submitted separately to 
several members of the experiment station staff, who were asked to 
classify them for color. The samples employed were: 

1. Preston, a red spring wheat. 

2. Turkey, a hard red winter wheat. 

3. Harvest King, a red wmter wheat common in the East, which 

under University Farm conditions is less corneous than 
Turkey. 

4. Kubanka, a durum wheat. 

5. Soft white wheat from the Pacific Coast. 

6. Preston selection which lacks red pigment in the bran layer. 

Three of the men who classified the samples used the term " amber " 
to describe color, and modified it by the use of such adjectives as 
"dark," ''light," "very light," etc. There was no consistent use of 
the defining adjective, however, and what one man termed "light 
amber " was called " dark amber " by others. A fourth classifica- 
tion used the term " amber " to express the appearance of corneous 
durum and other wheats which lack red pigment, and modified it by 
the use of the term " red " to denote a corneous wheat with red pig- 
• ment. The objection to this method is the use of a single term to 
express lack of bran pigment combined with a corneous endosperm. 
Three other men classified the samples by the use of two columns, 
one for pigmentation, the other for density of endosperm. There 
was no uniformity in the use of terms on the part of these men, 
however. This test indicated the need of a uniform method of color 
classification, and prompted the appointment of the committee which 
renders this report. 

It is generally recognized that the presence or absence of a red 
pigment in the bran layer is of little importance in indicating milling 
value, since there are certain unpigmented varieties which possess 
as good milling qualities as many of the pigmented varieties. The 
physical condition or density of the endosperm is of considerable 
iinportance because of its relation to milling properties and bread- 
making qualities. The relative stabihty of these two characters is 
of significance in a discussion of color classification. The first of 
these characters, pigmentation, is definitely inherited, and exhibits 
itself under widely varying conditions of environment. Although 
modified to some extent by climatic conditions the intensity of pig- 



COX.OR CLASSIFICATION OF W J I FAT. 



283 



mentation is a varietal character as well, some varieties possessing a 
less deoree of pignienlalion than others. With the same degree of 
pigmentation, a starchy kernel possesses a lighter appearance than a 
corneons kernel, however, Init there is no difficulty in separating 
a starchy pigmented kernel from a starchy white kernel. 

If we consider that inheritance is a characteristic manner of re- 
acting to a certain environment, we may say that the physical condi- 
tion, whether corneous or starchy, is an inherited character. Unlike 
the pigment, however, the density of the endosperm is very de- 
pendent upon environmental conditions. Thus, at University Farm, 
no wheats would be constantly starchy, although some varieties are 
consistently softer than others. Even though this character is easily 
influenced by environmental conditions, it is of such importance, be- 
cause of its relation to milling qualities, that a classification under 
this head is necessary for the breeder, farm crops expert, or milling 
chemist. 

The following scheme of classification is accordingly proposed. 
Columns headed (i) Pigmentation, and (2) Physical condition or 
density are necessitated. Under pigmentation the use of the term 
" red " to denote the presence of a brownish-red pigment in the bran 
layer is proposed. This is to be modified by the term " light " when 
the degree of pigmentation is less than is usual in red wheats. While 
the pigment may not be entirely absent from the bran layer of the 
so-called " white wheats," it is so nearly so that the term " white " is 
proposed in classifying them. It is recognized that a corneous kernel 
with a nonpigmented bran layer will not appear to be perfectly white ; 
so far as the color of the bran layer is concerned, it is not afifected 
by the density of the endosperm, however, although the complex 
of visual appearance due to the two factors is influenced by the 
relative endosperm density. 

Under physical condition or density it is proposed that four terms 
be employed to denote the several gradations of endosperm density. 
These are (i) corneous, (2) subcorneous, (3) substarchy, and (4) 
starchy. Under group i, corneous, w^ould be included only the uni- 
formly corneous samples. Group 2, subcorneous, would include 
samples containing kernels which approach either of the following 
conditions or a combination of both, — (a) samples containing % 
corneous kernels and ^ starchy or substarchy and (b) samples in 
which nearly all kernels approach the corneous group, the greater 
part of the kernels having only a small percentage of starchy endo- 
sperm. Group 3, substarchy, consists of kernels ^ of which are 



284 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



starchy and Ys corneous or kernels which contain a small amount 
of corneous matter with the larger part of the endosperm starchy, 
or a combination of these two conditions. Group 4 would be lim- 
ited to the uniformly starchy material. It is recognized that a 
sample will often be found to be intermediate between two of these 
groups ; in such cases it must be classified in the group which it most 
nearly resembles. The plus ( + ) and minus ( — ) signs may be em- 
ployed to designate that it varies somewhat above or below the aver- 
age of the group to which it is assigned in point of relative density. 

The application of the proposed system to the six wheat samples 
mentioned above would result as follows : 



Sample. 
I 



Pigmentation, 



Light red 
Red 

Light red 



Physical condition or density. 



Subcorneous 
Corneous 
Substarchy 
Corneous 
Starchy 
Subcorneous 



2 



3 
4 
5 
6 



White 
White 
White 



Minnesota Section, 
American Society of Agronomy. 



kelley: action of magnesium carbonate. 



285 



THE ACTION OF PRECIPITATED MAGNESIUM CARBONATE 

ON SOILS.i 

W. P. Kelley. 

Introduction. 

During recent years many investigations have been made on the 
lime-magnesia ratio in relation to plant growth. Different aspects 
of the question have been studied and many interesting experimental 
data obtained. In a large part of this work the attention of the 
investigator has been given mainly to the growth and yield of the 
crops employed. Just as in the study of many other soil problems, 
the final effects produced on the growing plant have been assumed 
to be brought about by the direct action of the substance applied. 
Consequently, attention has been focused on the physiological response 
on the part of the higher plant with only an occasional inquiry into 
the factors lying between the application and the response noted. It 
is obvious, therefore, that in such cases the soil has been looked upon 
as being in a state of stable equiHbrium. 

However, it is well known that, while soils without artificial treat- 
ment tend to undergo more or less continual change, the application 
of various substances, both organic and. inorganic, frequently induces 
various reactions in soils. And while knowledge regarding the 
specific nature of such reactions is very imperfect, sufficient is known 
to justify the statement that artificial applications produce changes 
of a greater or less degree in the chemistry, physics, and microbiology 
of soils generally, and that some, at least, of these changes can 
scarcely fail to reflect themselves on the physiological behavior of 
growing crops. 

It has been recognized for some time that the addition of soluble 
salts produces interchange of bases in soils, that the interchange when 
salts of different cations are added is not necessarily molecularly 
proportionate, and that it is rarely safe to predict from observations 
with one soil concerning the extent of interchange that will take 
place in another, since the rate of double decomposition in soils is 
quite variable. Nevertheless, the results of various experiments in-* 

^ Paper No. 43, University of California, Citrus Experiment Station, River- 
side, Cal. Received for publication July 2, 1917. 



286 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

volving the use of soluble salts have frequently been discussed by 
writers on the subject as if the soil were incapable of being reacted 
upon by the substances applied. Many references bearing on this 
point can be cited. Among such may be mentioned certain studies 
with the use of soils from the semiarid west on the effects produced 
by the addition of various soluble salts (3, 10, 11, 12). ^ 

Consideration of the principles of physical chemistry suggests, also, 
that other points of chemical equilibrium must necessarily be shifted 
as a result of the addition of soluble substances to soils. For 
example, it has recently been demonstrated in this laboratory that 
sodium carbonate is readily formed, in accordance with the prin- 
ciples of mass action, when sodium nitrate or other sodium salts 
are added to certain semiarid soils that contain calcium carbonate. 
In view of the ready solubility of sodium carbonate, the reaction of 
the soil solution must likewise be affected. Of course the reaction 
of soils may also be affected by the direct application of alkaline 
substances, such as carbonates, but in general these aspects of soils 
have not been sufficiently recognized in studies involving the applica- 
tion of chemical substances. Especially is this true in the investiga- 
tion on the lime-magnesia ratio. 

A number of compounds of calcium and magnesium have been 
employed in studies on the lime-magnesia ratio. One of the widely 
used compounds is the precipitated carbonate of magnesium. In cer- 
tain cases the effects produced by this substance have been quite 
unlike those following the application of other magnesium com- 
pounds. In 1904, Meyer (15) recorded the results of pot experi- 
ments with the use of chloride, sulfate, citrate, and carbonate of 
magnesium and calcium carbonate. The experiments were conducted 
in a sandy loam soil with mustard as the test crop. The results were 
very striking, as is shown by the following data. The total yield 
from three pots without the application of any substance was 43.5 
grams; with magnesium chloride, 40.5 grams; with magnesium sulfate, 
50.1 grams; with magnesium citrate, 150.5 grams; with magnesium 
carbonate, 146.0 grams ; and with calcium carbonate, 139.5 grams. 
These data show that, while the chloride and sulfate of magnesium 
produced only slight effects, the citrate and carbonate and also the 
calcium carbonate produced very striking stimulation. 

In still other experiments, Meyer found that the use of relatively 
large quantities of magnesium carbonate resulted in reduced yields 

2 Numibers in parentheses refer to papers similarly numbered in the bibli- 
ography on p. 295. 



KKLLKV: ACTION OF MAGNESIUM CARBONATK. 28/ 

in the case of rye, oats, lupines, peas, and carrots. In some cases a 
given (iiiantity of niai^nesiuni carbonate prevented growth altogether, 
while smaller quantities produced stimulation. In general Meyer 
found that the simultaneous addition of calcium carbonate and mag-^ 
nesium carbonate produced much the same effect as either when 
applied alone. It is probable that the results obtained from the use 
of magnesium carbonate and citrate and of calcium carbonate were 
due in considerable part to effects on the reaction of the soil rather 
than to modification of the lime-magnesia ratio. 

Hopkins (4) likewise found that while the application of magne- 
sium carbonate in quantities up to 0.8 percent of the soil produced 
notable stimulation in the yield of wheat in pot cultures, the addi- 
tion of 1.2 percent or more produced marked diminution in yield. 
In some cases complete failure of the crop was reported. These 
results likewise were probably due to changes in the reaction of the 
soil rather than to changes in the lime-magnesia ratio, as has already 
been pointed out by Gile (2). 

]\Iany other citations could be made to the use of magnesium car- 
bonate, the results of which are susceptible of the same interpreta- 
tion. In fact, the literature^ on this subject contains many references 
to experiments from various parts of the world in which it has been 
found that magnesium carbonate has produced injurious effects. 

In this connection it may be mentioned that burnt lime high in mag- 
nesia has long been looked upon as being likely to produce injury to 
vegetation. The older writings on agriculture (17) contain frequent 
reference to this fact. In general it has been assumed that the in- 
jury in such cases was due to excessive alkalinity occasioned by the 
magnesia. With few exceptions it appears, however, that excessive 
alkalinity has not been seriously considered in connection with the 
injury produced by precipitated magnesium carbonate. 

In 1912 the writer (5) began the study of the lime-magnesia ratio in 
relation to the biochemical formation of ammonia and nitrate in soils. 
The precipitated carbonates were employed as sources of calcium and 
magnesium and dried blood as the source of nitrogen. Two types of 
sandy soil from California were used. Later a more extended inves- 
tigation of the same general nature was made with a considerable 
range of soils from the Hawaiian Islands (6, 7). 

3 The reader is referred to a very complete bibliography of this subject by 
C. B. Lipman, Plant World, v. 19, no. 4, p. 83-115, and no. 5, p. 119-131, 1916. 
In the latter of these papers, Professor Lipman also makes reference to some 
original experiments in California in which magnesium carbonate has proved 
to 'be injurious to plant growth. 



I 



288 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

In these investigations the methods that have been widely used 
among American bacteriologists were employed. Briefly, the method 
as used consisted in incubating portions of soils with varying amounts 
of magnesium and calcium carbonate, after first having added i to 2 
percent of nitrogenous substance (dried blood, soy bean cake meal, 
or ammonium sulfate) and sufficient sterile water to produce suitable 
moisture content. After fixed periods of seven and twenty-one days, 
respectively, the ammonia and nitrates were determined. The re- 
sults thus obtained formed the basis of considerable discussion. 

It is not necessary at this time to discuss the several details of the 
method used in these studies, since the writer has already pointed 
out (8, 9) some of its irrational features and further discussion will 
be made elsewhere. Suffice it to say that after considerable study 
and investigation the writer believes that certain features of this 
method, especially the concentration of the nitrogenous substances 
employed, are so abnormal as to cast serious doubt on the practical 
value of the data obtained. 

At any rate, it is a matter of interest to note that under the con- 
ditions employed in the experiments referred to above, magnesium 
carbonate was found to be notably toxic to nitrification in certain soils 
and only slightly so in others. Calcium carbonate, on the other hand, 
was stimulating in certain soils and produced only slight effects in 
others. In regard to ammonification, magnesium carbonate was 
again found to be toxic when added to certain sandy or silty types 
of soil, but stimulating in the heavier types of soil where calcium car- 
bonate produced only slight effects. 

In no case did the further addition of calcium carbonate greatly 
modify the effects produced by magnesium carbonate alone. In con- 
trast to the effects of precipitated magnesium carbonate, the naturally 
occurring double carbonate of magnesium and calcium, dolomite, was 
found to produce effects similar in every way to that of calcium car- 
bonate, and was not toxic in any case, even in soils where precipitated 
magnesium carbonate was extremely toxic. In one soil, for example, 
with all other conditions the same, it was found that nitric nitrogen 
was formed at the rate of 670 p.p.m. where calcium carbonate was 
added, 700 p.p.m. where dolomite was added, and only 38 p.p.m. 
where the same percentage (2 percent) of precipitated magnesium 
carbonate was added. In another case, the addition of only o.i 
percent precipitated magnesium carbonate to a sandy type of soil 
from California almost completely suppressed the formation of nitrate, 
while the addition of calcium carbonate in quantities up to 8 percent 
produced notable stimulation. 



kelley: action of magnesium carbonate. 289 

In discussing these data (6, 7) the writer made some reference to 
the possibihty of excessive alkahnity having been a Hmiting factor 
where precipitated magnesium carbonate was used, but, on the whole, 
the view was expressed that the concentration of magnesium in the 
soil moisture was probably the more important factor. It was sug- 
gested that the concentration of magnesium in solution in soils might 
possibly become too high in certain cases for the progress of suitable 
biochemical action. But the results were interpreted as indicating 
that the ratio of calcium to magnesium is without significance so far 
as nitrification and ammonification are concerned. In the light of 
data presented below, however, it now seems more probable that the 
extreme toxicity produced by precipitated magnesium carbonate was 
due to excessive alkalinity. 

Lipman and Burgess (13) have more recently studied the effects 
of magnesium carbonate on nitrogen fixation by pure cultures of 
Asotobacter chroococcum and have found it to be distinctly toxic 
when added in more than very low concentrations. They likewise 
interpreted their results as being not due to excessive alkalinity but 
rather to excessive concentration of magnesium brought about by the 
treatment. 

In the course of some recent studies on the effects of various sub- 
stances on nitrification, data have been obtained which are of in- 
terest in this connection. 

Experimental Results, 

The experiments were made with the use of two light sandy loam 
soils low in organic matter, drawn near Riverside, Cal. Varying 
amounts of different substances (Baker's analyzed chemicals) were 
mixed with 100-gram portions of air-dried soil. After adjusting the 
moisture and incubating at 25° C. for four weeks, the nitrate was de- 
termined by the phenol-disulphonic acid method. In Table i are 
given the results showing the comparative effects of magnesium 
sulfate and magnesium carbonate on the nitrification of ammonium 
sulfate. The experiments were made in duplicate with closely agree- 
ing results. 

These data are in harmony with the results previously reported in 
showing that the addition of the precipitated magnesium carbonate 
may interfere with the formation of nitrate to a marked degree. 
As small an amount as 0.05 percent magnesium carbonate reduced the 
yield of nitrate in one soil from 84 p. p.m. to 70 p.p.m., and in an- 
other from 170 p.p.m. to 124 p.p.m., while the higher percentages pro- 
duced still more marked effects. Almost the same effects were pro- 



290 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

duced by o.i percent as by 0.5 percent in each soil. This may be ex- 
plainable by the fact that, since magnesium carbonate is a difficultly 
soluble substance, either of these amounts was sufficient to produce a 
saturated solution in the soil moisture. If so, the same effects might 
reasonably be expected to be produced in each case. 

On the other hand, the addition of magnesium sulfate up to as 
much as 0.5 percent produced no effect on nitrate formation. The 
effects of precipitated magnesium carbonate and magnesium sulfate, 
therefore, were widely different. 



Table i. — Comparative effects of magnesium carbonate and magnesium sulfate 

on nitrification. 



Materials added. 


Parts per million of nitric nitrogen. 








Soil No. I. 


Soil No. 2. 




10. 


34-5 


0.15 percent (NH40H)2S04 


84.0 


170.0 


0.15 percent (NH40H)2S04 and 0.05 percent MgCOs 


70.0 


124.0 


0.15 percent (NH40H)2S04 and o.io percent MgCOs 


17.0 


47.0 


0.15 percent (NH40H)2S04 and 0.50 percent MgCOs 


13-2 


44.0 


0.15 percent (NH40H)2S04 and 0.05 percent MgS04 


82.5 




0.15 percent (NH40H)2S04 and o.io percent MgS04 


83-5 




0.15 percent (NH40H)2S04 and 0.50 percent MgS04 


82.0 






1.2 


5-0 



It has been pointed out elsewhere (5, 9) that precipitated mag- 
nesium carbonate may interfere with the formation of nitrate with- 
out being toxic to the formation of nitrite. In 1915 Panganiban (16) 
also noted this fact in some studies on soils from the Philippine 
Islands. Since nitrite becomes reduced to ammonia under the in- 
fluence of the reducing agents employed in the several reduction 
methods that are in use for the determination of nitrate, it is evident 
that the amounts of nitrogen found by these methods should not be 
looked upon as being derived solely from nitrate unless actual test 
proves the absence of nitrite. 

The above results would seem to indicate that the marked effects 
of precipitated magnesium carbonate on nitrification previously noted 
were not so much due to modifications in the lime-magnesia ratio or 
to increases in the concentration of magnesium as to effects on the 
reaction of the soil. 

In further study of this subject, a considerable range of substances 
was used in order to obtain evidence regarding the importance of an 
interchange of bases as a factor of influence. Soil No. i was used, 
with the same concentration of ammonium sulfate as in the pre- 
ceding series. The results are shown in Table 2. 



kkijj:v: action ok maonksium carhonatk. 291 

Tarlk 2. — Coiiil^aralirr I'lfi'cls of diffrrcut snbslanccs on nitrification. 

Nitri<: 
nilrogtn, 

M atci i.ils added. jj.p.iii. 

None lo.o 

0.15 percent (NH,OH),SC)^ 89.0 

0.15 percent (NH.OH )oS04 and 0.05 percent Na,C"(), S6.o 

0.15 percent (NH40.H).,S04 and o.io percent Na.COy 33-5 

0.15 percent (NH40H)2SO., and 0.50 percent Na.COs i.i 

0.15 percent (NH40H)2S04 and 0.05 percent Na.S04 76.0 

0.15 percent (NH40H)2S04 and o.io percent Na.S04 60.0 

0.15 percent (NH40H)oS04 and 0.50 percent NaoSOi 41.0 

0.15 percent (NH40H)oS04 and 0.05 percent K2CO3 30.0 

0.15 percent (NH40H)2S04 and o.io percent K.COs i7-5 

0.15 percent (NH40H)oS04 and 0.50 percent K2CO3 2.0 

0.15 percent (NH40H)2S04 and 0.05 percent K2SO4 83.5, 

0.15 percent (NH40H)oS04 and o.io percent K2SO4 83.5. 

0.15 percent (NH40H)2S04 and 0.50 percent K2SO4 52.O' 

0.15 percent (NH40H)2S04 and 0.05 percent CaO 15.0 

0.15 percent (NH40H)2S04 and o.io percent CaO 6.6 

0.15 percent (NH40H)2S04 and 0.50 percent CaO 0.8 

0.15 percent (NH40H)2S04 and 0.05 percent CaSOi 84.0. 

0.15 percent (NH40H)2S04 and o.io percent CaSOi 87.5; 

0.15 percent (NH40H)2S04 and 0.50 percent CaSOi 84.0^ 

0.15 percent (NH40H)2S04 and 0.05 percent MgCOs 57.0- 

o.is percent (NH40H)2S04 and o.io percent MgCOs 17.0 

0.15 percent (NH40H)2S04 and 0.50 percent MgCOs 16.4 

0.15 percent (NH40H)2S04 and 0.05 percent MgS04 82.0 

0.15 percent (NH40H)2S04 and o.io percent MgSOi 85.0 

0.15 percent (NH40H)2S04 and 0.50 percent MgSOi 80.0 



The results shown in Table 2 indicate again that the depressing 
effect of precipitated magnesium carbonate on nitrification is due to 
the fact that this substance may produce excessive alkalinity in soils. 
It is shown, for example, that both sodium carbonate and potassium 
carbonate produced effects quite similar to magnesium carbonate,, 
while calcium oxide was somewhat more toxic than either of these 
substances. In view of the fact that calcium hydrate is known to be 
more strongly basic than magnesium carbonate, the results obtained 
from the use of calcium oxide may be interpreted as supporting the 
view that the toxicity of precipitated magnesium carbonate was due 
to injurious alkalinity. 

The results of the application of different sulfates also indicate 
that the differences in the effects produced by magnesium sulfate 
and magnesium carbonate can hardly be attributed to an interchange 
of bases in the soil. 

These data are also of interest in relation to the lime-magnesia 



292 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



ratio. The ef¥ects of increasing amounts of calcium sulfate were 
almost identical with those of magnesium sulfate, neither having 
produced any notable effect. As will be shown elsewhere, magnesium 
sulfate has the power of replacing small amounts of calcium from 
this soil, but only limited amounts. Chemical studies on this soil 
afford the proof that the ratio of calcium to magnesium in solution 
resulting from the application of calcium sulfate and magnesium 
sulfate must necessarily have been markedly altered by the amounts 
used above. Nevertheless, almost no effects were produced on the 
nitrification of ammonium sulfate. The conclusion seems warranted, 
therefore, that the nitrifying organisms have the power of adapting 
themselves to and can function equally well in the presence of con- 
siderable variations in the lime-magnesia ratio. 

The writer has pointed out elsewhere (9) that the effects of cer- 
tain soluble substances on the formation of nitrates differ widely when 
different concentrations of nitrogenous substances are employed. 
For example, with the use of i percent dried blood in one soil, 0.05 
percent sodium carbonate caused a diminution in the yield of nitrate 
from 172 p.p.m. to 31 p.p.m., while as much as 0.4 percent sodium 
carbonate produced no effect on the nitrification of o.i percent dried 
Wood. Likewise, in another soil, when 0.15 percent ammonium 
-sulfate was used, the addition of o.i percent sodium carbonate dimin- 
ished the yield of nitric nitrogen from 89 p.p.m. to 33.5 p.p.m. while 
marked stimulation resulted from the addition of the same amount of 
sodium carbonate when 0.0625 percent ammonium sulfate was used. 
Similar results have been obtained with the use of precipitated mag- 
nesium carbonate, as are shown in Table 3. 

Table 3. — The effects of magnesium carbonate on the nitrification of different 
concentrations of dried blood.<^ 



Parts per million of nitric nitrogen. 



Material added. 



None 

0.05 percent MgCOs 
o.io percent MgCOa 
0.50 percent MgCOa 

o Soil No. 2. 




102.0 
104.0 



These data show that marked diminution in the yield of nitrates 
from I percent dried blood resulted from the use of both 0.1 percent 
and 0.5 percent magnesium carbonate, while no effects were noted on 
the nitrification of 0.1 percent dried blood. 



KKLLEV: ACTION OF MA(;Ni:SlUM CAKHON ATIC. 293 

It is highly probable that the alkalinity due to the ammonia formed 
froni I percent dried blood became so high as to approach the toxic 
limit, in conseciuence of which only slight amounts of any other 
alkaline substance (magnesium carbonate) were rccjuircd to efifect 
prohibitive concentrations of alkalinity. 

Discussion. 

The preceding data show quite clearly that the effects produced by 
precipitated magnesium carbonate may dififer widely from those of 
magnesium sulfate. The addition of comparatively small amounts 
of the former retarded the formation of nitrate to a marked degree, 
while as much as 0.5 percent of the latter produced no effect. It was 
also shown that the addition of other alkaline reacting substances 
such as sodium and potassium carbonates and calcium oxide produced 
effects similar to magnesium carbonate. In view of the fact that mag- 
nesium sulfate produced no effect on the nitrifying process and that 
precipitated magnesium carbonate and dolomite have previously been 
found to produce widely different effects, the former being toxic and 
the latter stimulating, and since much of the ordinary precipitated 
magnesium carbonate is known to contain magnesium hydrate, the 
conclusion would seem to be justified that the toxic effects that have 
frequently been noted in studies with the use of this material have 
been occasioned by excessive alkalinity. The writer recognizes, how- 
ever, that it will be necessary to show that the toxicity of precipitated 
magnesium carbonate is positively correlated with the hydroxyl ion 
concentration before the above conclusion can be definitely drawn. 
Investigations are being made with this end in view. 

In any event, the data submitted above justify the conclusion that 
the inhibiting influence of precipitated magnesium carbonate toward 
nitrification was not due simply to excessive concentrations of the 
magnesium ion, for there can be little doubt that the larger amounts 
of magnesium sulfate produced considerably higher concentrations in 
the soil moisture than did the magnesium carbonate. 

In connection with the investigations of Lipman and Burgess (13) 
on the effects of magnesium carbonate on introgen fixation, referred 
to above, they sought to determine whether alkalinity was an im- 
portant factor. The method used consisted first in determining the 
degree of alkalinity of a saturated solution of the magnesium carbo- 
nate and then in studying the nitrogen-fixing power of Azotobacter 
in Ashby's mannite nutrient solution to which sufficient potassium 
hydrate had been added to effect a degree of alkalinity corresponding 
to that of a saturated solution of magnesium carbonate. They found, 



294 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

hpwever, that nitrogen fixation was stimulated by the addition of the 
potassium hydrate. Consequently the conclusion was drawn that the 
toxicity of magnesium carbonate was not due to injurious alkalinity, 
but to the magnesium ion. They say : 

The causticity of chemically pure magnesium carbonate as prepared by the 
Baker Chemical Company does not account for the toxic effects of magnesium 
carbonate. The latter effects must be due to the magnesium ion. The alkalinity 
of magnesium carbonate is beneficial, rather than otherwise, to A. chroococcum. 

Careful study of the conditions that ensued in the experiments of 
Lipman and Burgess will show, however, that the effects noted may 
still be interpreted as having been due in part, at least, to excessive 
alkahnity. For, upon the addition of potassium hydrate to a portion 
of Ashby's mannite solution, it is probable that slight precipitation 
of calcium and magnesium phosphates took place, thus introducing 
the solid phase, the absorptive power of which may have been a factor. 
But what is more important, the organic compounds that were prob- 
ably formed as a result of the decomposition of the mannite may 
have combined with the potassium hydrate and thus lowered the 
' alkalinity. 

Where magnesium carbonate was used, the smallest amount was in 
excess of that capable of being completely dissolved in the solution 
present. Consequently, potential alkalinity was present throughout 
the experiment in sufficient amounts to maintain a saturated solution, 
whereas the mannite solution containing potassium hydrate probably 
became less and less alkaline as the growth of the organisms pro- 
ceeded. 

It should also be mentioned that Ashby (i) found magnesium car- 
bonate to produce greater stimulation in nitrogen fixation by Azoto- 
bacter in mixed cultures than calcium carbonate, and in each case a 
greater amount was fixed than in neutral solutions. A full explana- 
tion of the discrepancy in his results, as compared with those of Lip- 
man and Burgess, can not be given. It is possible, however, that the 
magnesium carbonate which he used contained less magnesium hydrate 
than that used by Lipman and Burgess, and also that the activity of 
other organisms present tended to neutralize the magnesium carbon- 
ate. Ashby calls attention to the fact that butyric odors were pro- 
duced in his neutral cultures, but not where magnesium carbonate 
was used. In commenting on this point, he says : 

During concentration the neutral cultures developed a strongly acid odour,' 
those with calcium carbonate a weaker one, and those with magnesium car-- 
bonate, alone or mixed with calcium carbonate, gave no odour. When mag- 



KELLEV: ACTION OF MAGNESIUM CARBONATE. 295 

nesiiim carbonate was present, development was greatly delayed, but the yield 
of nitrogen was again larger, though not to so marked an extent as in the 
earlier experiment. In pure culture, Azolobacter gives rise to no acidity, either 
in solutions or on agar. One must conclude, therefore, that magnesium car- 
bonate not only neutralizes more effectually than calcium carbonate any trace 
of acidity due to foreign organisms in the early stages of culture, but also 
prevents butyric fermentation, but at first it inhibits the growth of Azotobacter 
itself. 

Finally, the investigations of IMcIntyre (14) afford strong evidence 
that precipitated magnesium carbonate is capable of exerting a 
strongly basic reaction in soils, as is shown by the fact that very 
large amounts of it were found to undergo decomposition with the 
evolution of carbon dioxide over long periods of time. 
, In view of the fact that the naturally occurring carbonates of 
magnesium produce widely different effects from the precipitated 
carbonate, together with the evidence set forth above, it seems reason- 
able to conclude that this material is unsuited for studies on the lime- 
magnesia ratio. With its use, effects on the reaction of the soil may 
so affect physiological processes as to obscure the effects that may be 
inherent within the ratio of calcium to magnesium itself, and there- 
fore the result obtained may lead to entirely erroneous conclusions. 

The data recorded above suggest the importance of the most 
thorough understanding of the chemical factors induced in soils by 
the application of treatments of different sorts. Soils are made up 
of complex yet varying mixtures of many substances, both organic 
and inorganic, the essential nature of some of which is known and of 
others little is definitely known. There is much evidence that the 
several points of chemical equilibrium in soils are easily shifted by 
the application of various substances, and the effects that finally mani- 
fest themselves on the physiological behavior of either the micro- 
organisms or of the higher plants may or may not be due to the direct 
action of the substance applied. 

Furthermore it is quite possible that the effects observed on the 
microbiological phenomena of soils do not bear a necessarily direct 
relation to the growth of crops. The urgent need in this as in other 
phases of soils and crop production is for exact knowledge concern- 
ing the chemistry and physics involved, without which the interpre- 
tation of soil biological phenomena must remain speculative. 

Literature Cited. 

I. ASHBY, S. F. 

1907. Some observations on the assimilation of atmospheric nitrogen by 
a free living soil organism, Azotohacter choococcum of Beije- 
rinck. Jour. Agr. Sci., v. 2, no. i, p. 35-51. 



296 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



2. GiLE, P. L. 

1912. Lime-magnesia ratio as influenced by concentration. Porto Rico 
Agr. Expt. Sta. Bui. 12, p. 8-9. 

3. Harris, Frank S. 

1915. Effect of alkali salts in soils on the germination and growth of 
crops. U. S. Dept. Agr., Jour. Agr. Research, v. 5, no. i, p. 
1-53. 

4. Hopkins, Cyril G. 

1910. Soil fertility and permanent agriculture, p. 170^171. Ginn & Co., 

Boston. 

5. Kelley, W. p. 

1912. The effects of calcium and magnesium carbonates on some bio- 
logical transformations of nitrogen in soils. Univ. of Cal. Pub. 
Agr. Sci., V. I, no. 3, p. 39-49- 

6. . 

1914. The lime magnesia ratio. I. The effects of calcium and mag- 

nesium carbonates on ammonification. H. The effects of calcium 
and magnesium carbonates on nitrification. Centbl. f. Bakt., 2 
Abt. 42, nos. 17-18, p. 519-526; nos. 21-23, p. 577-582. 

7- . 

1915. Ammonification and nitrification in Hawaiian soils. Hawaii Agr. 

Expt. Sta. Bui. 37, p. 1-52. 

8. . 

1916. Some suggestions on methods for the study of nitrification. Sci- 

ence, n. s.^ V. 43, no. 1097, p. 30-33. 

9. . 

1916. Nitrification in semiarid soils — I. U. S. Dept. Agr., Jour. Agr. 
Research, v. 7, no. 10, p. 417-437 

10. LiPMAN, C. B. 

1911. Toxic effects of alkali salts in soils on soil bacteria. I. Ammoni- 

fication. Centbl. f. Bakt., 2 Abt. 32, no. 1-2, p. 58-64. 

11. . 

1912. Toxic effects of alkali salts in soils on soil bacteria. H. Nitrifica- 

tion. Centlbl. f. Bakt., 2 Abt. 33, no. 11-14, P- 305-313- 

12. . 

1914. Antagonism between anions as affecting soil bacteria. Centbl. f. 
Bakt, 2 Abt. 41, no. 11-17, p. 430-444. 

13. , and Burgess, P. S.. 

1914. The protective action against MgCO^ of CaCO^ for A. 
chroococcum. Jour. Agr. Sci., v. 6, pt. 4, p. 484-498. 

14. MacIntire, W. H., Willis, L. G., and Hardy, J. I. 

1914. The nonexistence of magnesium carbonate in humid soils. Tenn. 

Agr. Expt. Sta. Bui. 107, p. 151-202. 

15. Meyer, D. 

1904. Untersuchungen iiber die Wirkung verschiedener Kalk und 
Magnesiaformen. Landw. Jahrb., 33 : 371-404. 

16. Panganiban, E. H. 

1915. A study of nitrification in Philippine soils. Philippine Agr. & 

Forester, v. 4, no. 4, p. 81-91. 



connkr: icxckss salts in humid soils. 297 
17. Storfr, F. H. 

iyo6. Agriculture in some of its relations with chemistry, (7th ed.) 2: 
516-521. (New York.) 



EXCESS SOLUBLE SALTS IN HUMID SOILS.^ 

S. D. C0NNER.2 

When any soil in Indiana or the adjoining States in fair physical 
condition fails to produce good crops the usual procedure is to try to 
find what is lacking, so that lime, organic matter, or some fertilizing 
element may be supplied in order that better crops may be grown. 
\Miile most soil troubles of a chemical nature in humid regions are 
due to lack of some necessary ingredient, there are some cases where 
the trouble is due to an excess of one or more soluble salts. The 
injury caused by these soluble salts may be due in some cases to toxic 
action of more or less dilute solutions of certain salts or it may be 
due to plasmolysis caused by relatively concentrated solutions of 
salts that are otherwise not injurious. 

The occurrence of excessive amounts of soluble salts in soils of 
humid regions has been reported by Cameron^ as occurring in Mary- 
land, Florida, etc. The Soils and Crops Department of the Indiana 
Agricultural Experiment Station* has reported unproductive soils 
which contained excessively high nitrate and other soluble salts. 
Stevenson and Brown^ have also noted such soils as occurring in 
Iowa. Ames and Schollenberger^ report Ohio soils containing large 
percentages of soluble salts. 

The following cases relate to humid soils containing rather con- 
centrated salt solutions. In September, 19 13, two samples of peat 
soil were received from a farm near Toto, Starke Co., Ind. The 

1 Contribution from Soils and Crops Department, Indiana Agricultural Ex- 
periment Station, Received for publication March 29, 1917. 

2 The writer desires to acknowledge the assistance of Mr. H. R. Smalley, 
who performed some of the analytical work presented in this paper. 

3 Cameron, F. K. Soil solutions. U. S. Dept. Agr., Bur. Soils Bui. 17, p. 36. 
1901. 

* Rept. Soils and Crops Dept., 26th Ann. Rept. Ind. Agr. Expt. Sta., p. 60. 
1913. 

5 Stevenson, W. H., and Brown, P. E. Iowa peat and alkali soils, Iowa Agr. 
Expt. Sta. Bui. 157. 1915. 

6 Ames, J. W,, and Schollenberger, C. J. Accumulations of salts in Ohio 
soil. In Soil Science, vol. i, no, 6, p. 575, 1916. 



298 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

soil from a part of the field where onions were doing well contained 
0.45 percent soluble salts with o. 10 percent NO3. The sample from 
a part of the field where onions were dying contained 1.2 percent sol- 
uble salts with 0.44 percent NO3. One week later samples from 
another part of the same field were taken at depths of i inch, 6 inches, 
and 18 inches. The determinations of soluble salts and of nitrates 
are given in Table i. 



Table i. — Percentages of soluble salts and of nitrates at different depths from 
two samples of peat soil from^ an onion field in Starke Co., Ind. 



Depth of soil sample. 


Onions good. 


Onions poor. 


Soluble salts. 


NO3. 


Soluble salts. 


NO3. 


to I inch 


0.22 


0.100 


0.52 


0.3400 




•13 


•045 


.22 


.0600 


6 to 18 inches 


.11 


.060 


.14 


.0025 



Samples taken in July, 1916, near Cromwell, Ind., on peat soil 
contained 1.12 percent of soluble salts and 0.50 percent NO3 where 
the onions were doing poorly and 0.57 percent of soluble salts and 0.17 
percent NO3 where the onions were still good. ,In a number of 
other instances where onions were doing poorly on peat soils high 
percentages of soluble salts have been found. 

The analyses of four unproductive black soils are given in Table 2 
and the composition of water extracts of the same soils is shown in 
Table 3. These four soils may be described as follows : 

A. Unproductive peat from Noble Co. This sail failed to produce corn or 
onions after fertilization with potash and phosphate. 

B. Unproductive peat from Starke Co. on which onions failed to grow after 
potash and phosphate fertilization. 

C. Unproductive peat from Kosciusko Co. This soil had never produced a 
crop before or after fertilization. 

D. Unproductive peaty sand from Wanatah Field in Laporte Co., Ind. 
Crops did not grow on this soil until 2 tons to the acre or more of limestone, 
together with acid phosphate and potash, were applied. Good crops were pro- 
duced after this treatment. 

Soils A and B are of a type similar to the Toto and Cromwell 
soils where at certain seasons of the year the concentration of salts 
at the surface may be easily high enough to plasmolyze the tissue of 
tefrider growing plants. These soils are often coated with a white 
crust and may be truly called alkali soils. The composition of the 
water extract as shown in Table 3 would indicate that the bulk of 
tjae- alkali salts are nitrates with lesser amounts of sulfates and 
chlorids. Calcium is the most abundant base. While these soils are 



CONNER: K\(i:SS SALTS IN HUMID SOUS. 299 



Table 2. — Analyses of four unproductive black soils in Indiana. 



Determination. 


Soil A, 


ouil n. 




Soil D 




Percent 




X Cr WfC ritf , 


PcYceyil. 




8.96 




14. / i 


88.63 




•17 






T A 




.14 


.18 


.06 




CaO 


332 


2.99 


•03 


.08 


MgO 


•43 


•52 




.11 


AT M ^0 , 


Trace 






Trace 




.72 


• / 




78 


AhOs 




1.28 


I 07 


2.64 


p,Qj. 








.08 


SO3 


.56 


.66 


.21 


.10 


CO2 


•32 


Trace 


Trace 


Trace 


Volatile 


79.41 


87.72 


83.53 


8.16 


Total humus 


33-58 pet. 


31.09 pet. 


44-45 pet. 


4.86 pet. 


Aeid humus'* 


10.06 pet. 


16.64 pet. 


43.08 pet. 


4.64 pet. 


Total nitrogen 


3-57 pet. 


4.03 pet. 


2.34 pet. 


.28 pet. 


Nitrates (NOs)'^ 


1,000 ppm. 


2,670 ppm. 


167 ppm. 


134 ppm. 


Aeidity'' 


250 lbs. 


750 lbs. 


8,000 lbs. 


3,500 lbs. 



« Digestion in HCL, sp. gr. 1.115. 

* Extracted without previous acid treatment. 

c Phenol di-sulfonic acid method. 

^ Potassium nitrate method. Results expressed in pounds of CaCOs per 
2,000,000 pounds of soil. 



somewhat acid in reaction, they contain a large amount of calcium 
and organic nitrogen. During the warm weather of spring and 
summer, active nitrification sets in and large amounts of nitrates 
accumulate. These nitrates, together with other soluble matter, are 
brought to the surface and are deposited when the soil moisture 
evaporates. This type of soil gives no trouble and shows no exces- 



Table 3. — Composition of the water extract from the four unproductive black 
soils, the analyses of which are given in Table 2. 



Determination. 


Soil A. 


Soil B. 


Soil C. 


Soil D. 




Grams. 


Grams. 


Grams. 


Grams. 


Si02 


"0.0720 


0.0312 


0.0078 


0.0161 


K 


.0456 


.1080 


•0321 


.0118 


Na 


.0393 


• 0737 


•0158 


.0342 


Ca 


.2070 


.6066 


.0113 


.0200 


Mg 


•0335 


• 1573 


.0072 


.0202 


Fe 


.0000 


.0002 


.0000 


.0010 


Al 


.0025 


.0270 


•0155 


.0172 


NH4 


.0053 


.0208 


.0667 


.0062 


PO4 


.0044 


.0145 


.0103 * 


.0080 


SO4 


.1543 


•3567 


.0855 


.0177 . 


NO3 


.7407 


2.6667 


.1066 


.2500 


CI 


.1332 


.3000 


.0613 


.0693 


Organic matter, etc 


.1762 


•3040 


.5482 


• 1577 



o Grams dissolved by leaching one kilo of soil with 4,000 cc. distilled water. 



3 CO JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

sive accumulation of salts when left in grass and not cultivated. 
Trampling by stock as in pasturing has a tendency to benefit them 
permanently. Rolling with very heavy power rollers has been tried 
with more or less success. The more compact they are kept the less 
active is the nitrification and the less the salts tend to accumulate. 

Soils C and D are types found in some localities in Indiana. Such 
soils are quite low in calcium or other bases and are very acid. The 
water extracts show relatively high concentrations of aluminum salts 
which have been shown to be quite toxic to plant roots in very dilute 
solutions. This type of soil does not show very high concentrations 
of total salts. There appears to be little doubt that the unproductive- 
ness of this type of soil is due principally to the presence of soluble 
aluminum salts. All very acid black soils which have been examined 
by the writer have been found to contain water soluble salts of 
aluminum. In some black soils doubtless the infertility may be due 
to a combination of high soluble salts and to toxic substances. 

There are also large tracts of peat and peaty sand soils in Indiana 
which do not come in either class of the soils which have been dis- 
cussed. Probably the largest areas of peaty and muck soils in this 
State are deficient in potash. There are other areas which respond 
only to phosphate.^ 

Other than the black soils, there are probably no alkali soils of 
natural origin in Indiana. Quite often, however, some land owner 
sends to the experiment station a sample of soil which will not pro- 
duce crops on account of too much soluble salts. Such soils are gen- 
erally of artificial origin caused by accumulation of refuse. A farmer 
near Warren, Ind., reported a spot of soil in a field on which nothing 
would grow. The percentages of water-soluble materials found in 
samples of this soil taken at depths of 0-6, 12-18, and 24-30 inches 
are shown in Table 4. 



Table 4. — Water-soluble materials in silt loam soil from unproductive spot near 

Warren, Ind. 



Determination. 


Depth of sample. 


0-6 inches. 


12-18 inches. 


24-30 inches. 




Percent. 


Percent. 


Percent. 


Nitrates (NO3) 


O.IO 


•03 


0.012 


Potash (KtO) 


■85 


.42 


.440 


Soluble salts 


2.54 


1.40 


1.330 



7 Abbott, J. B., Conner, S. D., and Smalley, H. R. The reclamation of an 
unproductive soil of the Kankakee marsh region. Ind. Agr. Expt. Sta. Bui. 170. 
1913. 

8 Conner, S. D., and Abbott, J. B. Unproductive black soils. Ind. Agr. Expt. 
Sta. Bui. 157. 1912. 



connkr: kxckss salts in humid soils. 301 

The analysis in Table 4 shows that the uni)rocluclivity was due to 
the presence of water-soluble salts, principally potash. This unj)ro- 
(luctive place was in a part of the field where a stable had been 
located up to five years previous. This is a good demonstration of 
the fact that plant food is being continually lost by leaching in stables 
that do not have water-tight floors. In this case the farmer was ad- 
vised to dig out the unproductive spot and spread it on the remainder 
of the field as manure. Similar soils from other parts of the State 
have been tested with like results. 

Summary. 

1. Black soils in humid regions sometimes contain excessive 
amounts of soluble salts. 

2. These soluble salts may cause injury to crops, due to high con- 
centration of nontoxic salts, to a lower concentration of more toxic 
substances, or to a combination of both. 

3. The salts occurring in high concentration are generally nitrates. 

4. The toxic salts occur generally in acid soils and are mainly 
soluble salts of aluminum. 

5. The only clay and loam soils that were found to contain exces- 
sive soluble salts were of artificial origin, such as spots where old 
stables had stood. 

Indiana Agr. Expt. Station, f 
La Fayette, Indiana. 



302 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



AGRONOMIC AFFAIRS. 



THE ANNUAL MEETING. 

The Executive Committee of the American Society of Agronomy, 
at a meeting last November, agreed to hold the next meeting at the 
same place as and on the two days preceding the meeting of the As- 
sociation of American Agricultural Colleges and Experiment Stations. 
At that time it was the plan of the latter organization to meet in 
Springfield, Mass., in October, in connection with the semicentennial 
celebration of the Massachusetts Agricultural College. This celebra- 
tion has been abandoned and the Executive Committee of the A. A. 
A. C. E. S. has decided to meet in Washington, D. C, about the 
middle of November. The tenth annual meeting of the American 
Society of Agronomy will be held, therefore, in Washington, pre- 
sumably on November 12 and 13. All those who expect to attend 
and to present papers are requested to send titles to the Secretary at 
the earliest possible date. Also, those who plan to come are asked to 
inform the Secretary as to whether or not they will attend an agrono- 
mists' dinner on Monday, November 12, if one is arranged. 

MEMBERSHIP CHANGES. 

The membership of the Society, as reported in the May issue, was 
643. Since that time 9 new members have been added, 4 have been 
reinstated, and i has resigned, a net gain of 12 and a present total 
membership of 655. The names and addresses of the new and rein- 
stated members and the name of the member who has resigned, with 
such changes of address as have come to the notice of the Secretary, 
are as follows : 

New Members. 

Daane, Adrian, 225 Duncan St., Stillwater, Okla. 

Fleming, Frank L., Jireh, Wyo. 

Graham, E. E., R. No. 2, Stonewall, Okla. 

Jarvis, Orin W., Pacific Sugar Corporation, Tracy, Gal. 

Kime, p. H, West Raleigh, N. C. 

Murphy, Henry, 318 West St., Stillwater, Okla. 

Riley, J. A., Chester, S. C. 

Spencer, E. L., 210 Elm St., Stillwater, Okla. 

Ware, J. O., West Raleigh, N. C. 



AC.KONOMIC Al- I AlKS. 



MKMin<:fiS RlilNSTATKl). 

Chapman, James E., 2316 Pierce Ave., St. Anthony Park, St. Paul, Minn. 
C'URREY, Hiram M., Bureau of Markets, U. S. Dept. Agr., Washington, D. C. 
Lechner, H. J., Washington State Normal School, Ellensburg, Wash. 
Wilson, Bruce S., Experiment Station, Manhattan, Kans. 

Member Resigned. 

Peters, David C. 

Addresses Changed. 

Barker, Joseph F., College of Agriculture, Columbus, Ohio. 
BoviNG, Paul A., University of British Columbia, Vancouver, B. C. 
Burnett, Grover, Mackay, Idaho. 

Currey, Hiram M., Bureau of Markets, U. S. Dept. Agr., Washington, D. C. 

Dorsey, Henry, Agr. Expt. Sta., Morgantown, W. Va. 

Fletcher, O. S., E. 915 Augusta Ave., Spokane, Wash. 

Hendry, Geo. W., University of California, Berkeley, Cal. 

Hill, Pope R., 215 Lucy Ave., Memphis, Tenn. 

HoDSON, Edgar A., Box 285, West Raleigh, N. C. 

Holland, B. B., 800 Jefferson St., Amarillo, Tex. 

Hulbert, Harold W., Farm Crops Dept., University of Idaho, Moscow, Idaho. 
Hutchison, C. B., Dept. Plant Breeding, Cornell Univ., Ithaca, N. Y, 
McAdams, James, care State Board of Agr., Topeka, Kans. 
Miller, Frank R., Room 2132, 2 Rector St., New York, N. Y. 
Miyake, Koji, College of Agr., Tohoku Imp. Univ., Sapporo, Japan. 
Schick, G. M., Plainview, Tex. 

Olson, M. E., Soils Section, Iowa State College, Ames, Iowa. 
Sleeth, E. C, Jefferson, Ohio. 
Van Evera, R., Abraham, Utah. 

Smith, Raymond S., 304 Elmwood Ave., Ithaca, N. Y. 
Taggart, J. G., Lower Onslow, Nova Scotia, Canada. 

NOTES AND NEWS. 

H. W. Barre has been appointed director of research at Clemson 
College, S. C. He will be director of the experiment station but not 
dean of the college ; the latter office will not be filled at present. 

N. Eric Bell, formerly with the Alabama State soil survey, is now 
county agent in Hale Co., Ala., with headquarters at Greensboro. 

Sidney Bliss of Ohio State IJniyersity has been appointed assistant 
in the division of soils at the Ohio station. 

W. C. Boardman and O. H. Smith, assistants in the soil survey 
at the Ohio station, have resigned. 

P. A. Boving, who has been in charge of root investigations at 
Macdonald College, is now assistant professor of agronomy in the 
University of British Columbia. 



304 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

A. D. Hall, formerly director of the Rothamsted station and well- 
known writer of agricultural books, has been appointed permanent 
secretary of the Board of Agriculture and Fisheries of Great Britain. 

J. G. Hamilton, assistant agronomist of the New Mexico station, 
resigned March i to become county agent of Valencia Co., N. Mex. 

E. C. Higbie, superintendent of the West Central School of Agri- 
culture and of the substation at Morris, Minn., has resigned, effective 
July 31. 

H. W. Hulbert, a graduate of the Michigan college and post- 
graduate of the Iowa college (191 7), has been appointed to the farm 
crops department of the University of Idaho. 

Dr. George E. Ladd has resigned as president of the New Mexico 
college and has been succeeded by Dr. A. D. Crile. 

C. E. Neff is assistant in agronomy at the Delaware college and 
station. 

M. E. Olson, scientific assistant in corn investigations in the 
U. S. Dept. of Agriculture for the past year, has returned to the 
Iowa college, where he will superintend cooperative experiments in 
soil fertility over the State. 

George Severance, formerly professor of agriculture of the Wash- 
ington college, is acting director of the Washington station, succeed- 
ing Ira D. Cardiff. 

V. M. Shoesmith, professor of farm crops and farm crops ex- 
perimentalist at the Michigan station since 1910, is now superin- 
tendent of a 4,300-acre tract of land near Grand Rapids, Mich. 

R. R. Spafford is now assistant in farm management at the 
Nebraska station. 

A section of the American Society of Agronomy has been estab- 
lished at the University of Illinois, with Robert Stewart president, E. 
A. White, vice-president, and E. A. Torgerson, secretary-treasurer. 

The State of Texas has recently located junior agricultural colleges 
at Stephenville and Arlington. At Stephenville the buildings and 
grounds of John Tarleton college have been acquired and 500 acres 
of land donated for experimental and demonstration purposes. A 
locating board has also inspected numerous sites with a view to locat- 
ing the new West Texas Agricultural and Mechanical College. 



JOURNAL 

OF THE 

American Society of Agronomy 



Vol. 9. October, 191 7. No. 7. 



THE EFFECT OF PROLONGED GROWING OF ALFALFA ON THE 
NITROGEN CONTENT OF THE SOIL.^ 

C. O. SWANSON. 

Grain crops get their nitrogen from the soil. The ultimate source 
of this nitrogen is the air. Over each acre there is enough nitrogen 
to produce 50,000,000 bushels of corn. All crops get their carbon 
from the air. Over each acre there is only enough carbon to produce 
200 bushels, yet carbon has never figured in the commercial valuation 
of plant food, while under normal conditions nitrogen is the most ex- 
pensive element. (Abnormal conditions due to the great war make 
potassium at present the most expensive.) The fundamental reason 
for this difference between nitrogen and carbon is that all green 
plants have their own physiological organ for obtaining carbon from 
the air, while they have no such corresponding organ for obtaining 
nitrogen. Thus, while plants are bathed in an atmosphere approxi- 
mately four fifths nitrogen, they starve for the want of it unless they 
can get it from the soil. On the other hand, they get all the carbon 
they need, although the atmosphere contains only about o.oi percent. 

The power of legumes to improve the crop-producing power of the 
soil was known to the ancients. Just how this improvement came 
about remained for the modern chemist and bacteriologist to demon- 
strate. The scientific facts connected with these phenomena are so 
well known that it is not necessary to dwell on them here, but, like 
all new discoveries, certain phases are likely to be overworked. Stu- 

1 Contribution from the Department of Chemistry of the Kansas Agricul- 
tural Experiment Station, Manhattan, Kans. Read at the spring meeting of 
the American Chemical Society, Kansas City, Mo., April, 1917. This is a par- 
tial report of work still in progress. Received for publication May 10, 1917. 

305 



306 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



dents of soil fertility who have studied deeply and who think clearly 
have questioned some of the enthusiastic statements made in regard 
to the power of legumes to restore and maintain the crop-producing 
power of the soil. 

No attempt is made in this paper to review the literature on the 
subject. From experimental work under scientific control as well as 
from the practical experience of farmers, there is abundant evidence 
as to the power of legumes to improve the crop-producing power of 
the soil. The only question of which there is any doubt is in regard 
to the amount of nitrogen added to the soil when legumes are grown 
under conditions usually found on the American farm. As com- 
pared with other legumes, particularly red clover, little work has been 
done with alfalfa. This is partly due to the fact that alfalfa has not 
been grown generally by farmers in this country for as long a time 
as red clover, and partly to the fact that alfalfa is not so readily 
adapted to rotations. When a field has a good stand of alfalfa the 
value of this crop is so great that it is often more profitable to con- 
tinue growing it there than to plow it under and grow something else. 

A few citations from reports of work on this phase of soil fer- 
tility illustrate the general situation. Knorr^ makes the following 
statement : In every case where crop followed alfalfa the highest 
average yields were obtained, indicating very strongly that the alfalfa 
has a beneficial effect on the succeeding crop." This is a usual con- 
clusion from experiments of this nature and such results are entirely 
possible without the actual increase in the stock of nitrogen in the 
soil. Lyon and BizzelP grew alfalfa and timothy for six years on 
adjoining plots. On plowing up these were planted to corn the first 
year and to oats the second. The yield of corn grain was 62 bushels 
on the alfalfa plot and 47 on the timothy plot. The oats yielded 26 
bushels on the alfalfa and 27 on the timothy plot. Analysis of the 
soils from the two plots showed that the alfalfa soil contained not to 
exceed o.oi percent more nitrogen than the timothy soil. This would 
amount to 250 pounds per acre if the soil is assumed to weigh 
2,500,000 pounds to a depth of 8 inches. Anyone who has sampled 
and analyzed soil knows the difficulty of working on such a small 
margin. This work of Lyon and Bizzell raises the two questions of 
most importance in connection with the relation of legumes to soil 
fertility: (i) Was there a greater accumulation of nitrogen in the 

2 Knorr, F. U. S. Dept. Agr., Bur. Plant Ind., Rept. of work on Scottsbluff 
experiment farm, 1913. 

3 Lyon, T. L., and Bizzell, J. A. Experiments concerning the top-dressing of 
timothy and alfalfa. N. Y. Cornell Agr. Expt. Sta. Bui. 339. 1913. 



I 



SWANSON : EKFIX T OF ALFALFA ON SOIL NITKOdKN. 307 

alfalfa soil (luriiii^- the six \cars than there was in the timothy soil 
during the same length of time? (2) Was the greater productivity 
of the alfalfa soil due to the greater availa'hility of the nitrogen? 

Alway and Bishops report analyses of samples of soil from an 
alfalfa field and from a corn held. They conclude that "no marked 
difference is to be seen between the amounts of nitrogen in the soil of 
the two fields as represented by the samples analyzed." 

This paper does not deal with the question of whether the stock 
of nitrogen in the soil can be increased through the growing of 
legumes (in this case, alfalfa). The question is: When alfalfa has 
been grown on land for a number of years and all the crop has been 
harvested as hay, no return of any kind having been made to the 
soil, has the stock of nitrogen been increased on that particular piece 
of land? 

To find out what prolonged growing of alfalfa does to the soil was 
a question taken up by the Kansas Agricultural Experiment Station 
tw^o years ago. A number of fields in Kansas have been continu- 
ously in alfalfa for twenty to thirty years or more. In most cases 
it is possible to find near by soil of the same type which has been con- 
tinuously under cultivation since it was broken, thirty-five or forty 
years ago, and soil in native sod, used either as pasture or as hay 
land. By sampling and analyzing the soil in the various fields it is 
possible to learn something in regard to the efifect of long produc- 
tion of alfalfa on the soil. This experiment was carried on in co- 
operation between the departments of chemistry and agronomy. The 
samples have all been taken by the author, either alone or with the 
assistance of field members of the Division of Extension.^ 

The samples were generally taken in four different strata, namely : 
0-7 inches ; 7-20 inches ; 20-40 inches ; and 40-80 inches. For the 
upper three strata a soil auger was generally used. For the lowest 
it was found best to use a soil tube, particularly in western Kansas. 
Borings were made in a number of places in the field sampled. In 
most cases the alfalfa field, the field in native sod, and the field under 
continuous cultivation were in very close proximity, separated only 
by a fence or a road. In western Kansas most of these fields were 
on bottom soil. Anyone who has sampled soil for chemical analysis 
knows that bottom soils are most difficult to sample because of the 

* Alway, F. J., and Bishop, E. S. Some notes on the alfalfa and clover resi- 
dues as sources of soil nitrogen. In 25th Ann. Rpt. Nebr. Agr. Expt. Sta., p. 
56-65. 191 1. 

^ To all who have assisted in this work the author wishes to express his 
appreciation, particularly to Mr. W. L. Latshaw, who is in charge of the chem- 
ical soil laboratory. 



308 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

abrupt changes that are found, especially in strata below the surface. 
Because of this difficulty the results of analysis do not always show 
the correlation expected. In this paper results on nitrogen only are 
presented. The samples were analyzed also for phosphorus, calcium, 
and carbon, the latter in order to obtain data on the amount of or- 
ganic matter. 

The results of the chemical analysis for nitrogen are given in 
Table i. This table gives the county where the sample was obtained, 
the sample number, the cropping condition of the field, and the per- 
centages of nitrogen found in the four different strata. In making a 
preliminary study of these figures it was found that the results could 
be classified under three heads, viz : samples taken in the eastern 
portion of the state, where the rainfall is about 30 inches or more; 
samples taken in the west central part, where the rainfall is less than 
30 but more than 20 inches, and samples taken west of the line of 
20-inch rainfall. These regions will be referred to as humid, semi- 
humid, and semiarid. 

One of the first facts noticed is that in the semiarid portions of the 
state the alfalfa soils have at least as high a percentage of nitrogen 
as the soils in native sod and that the difference between the culti- 
vated soils and the soils in native sod is small. In humid regions all 
the soils in native sod contain a larger percentage of nitrogen than 
the soils in alfalfa, while with a few exceptions the soils in alfalfa 
contain a larger percentage of nitrogen than those continuously 
cropped. It should be noted that the alfalfa fields in the humid 
section were on the average less than two thirds the age of the fields 
in the semiarid section. 

In the semihumid section the results resemble both the semiarid 
and the humid. In some cases the alfalfa soil has more nitrogen 
than the native sod, and in other cases the native sod has the more. 
The fields here were somewhat older than in the humid section, but 
on the average not as old as those in the semiarid. 

Thus far the comparison is direct and simple. When the attempt 
is made to figure how much nitrogen has been stored in these alfalfa 
soils the problem becomes complicated. The nitrogen content of the 
fields in alfalfa at the time when these soils were first seeded is not 
known. All had been in cultivation for some time ; this time was 
longest in the humid section and shortest in the semiarid. As soon 
as native sod is broken up and cultivation commences, loss of nitro- 
gen begins. Some of the nitrogen is removed by the crop, but, as is 
well known to students of soil chemistry, the greater relative loss in 
new soil is through decomposition and oxidation of the organic mat- 



SWANSON : EFKl'XT OV ALFALFA ON SOIL NITKOCKN. 



Table i, — I'crcoitiu/c of nitrogen in different strata of some Kansas soils. 

HUMID SECTION. 



County. 



Soil 
No. 



Description of soil. 



Treatment. 



Character. 



Depth of sample. 



0-7 7-20 20-40 40-80 



Brown , 
Do. , 

Do. , 



Nemaha. 
Do. . 
Do. . 



Leavenworth. 
Do. 

Montgomery 
Do. 
Do. 

Do. 
Do. 

Dickinson . . . 
Do. ... 
Do. ... 

Do. 
Do. 
Do. 



1768 AhaUa 28 years 

1770 Native pasture, white 
clover 

1769 Cropped to grain 45 j'cars 

'1765 Alfalfa 21 years 
1767 Native pasture, blue stem 
1 1 766 Cropped to grain 45 years 

1318 Alfalfa 14 years 

1319 Native meadow, blue stem 

1294 Alfalfa 12 years 

1296 Native meadow, blue stem 

1295 Cropped to grain 35 years 

1297 Alfalfa 10 years 

1298 Cropped to grain 40 years 

1874 Alfalfa 20 years 

1876 Native meadow, blue stem 

1875 Cultivated, mostly corn, 

35 years 

1877 Alfalfa 20 years 

1879; Native pasture, blue stem 

1878 Cultivated, grains, 40 
I years (manured) 



Rolling upland 
D6. 

Do. 

Do. 
Do. 
Do. 

Do. 
Do. 

Upland 
Do. 
Do. 

Do. 
Do. 

Bottom 
Do. 
Do. 



Upland 
Do. 
Do. 



0.21 1 
.229 

.160 

.171 
.181 
.160 

.222 
.296 

.131 
.186 
.110 

.168 
•135 

.168 
.204 
.140 



.179 
.204 
.163 



0.142 
•133 

•135 

.081 
.085 
•135 

.177 
.225 

.106 
.114 
.094 

.087 
.076 

.117 
•134 
.101 



•134 
■ 131 
.120 



0.062 0.031 
.073 -048 

.066 .031 

.032 I .008 
.028 .008 
.066 1 .031 



.121 
.144 

•054 
.067 
.050 

.038 
.043 

.084 
.109 
.084 



.077 
.080 
.070 



.082 

.031 
.024 
•039 

.028 
.026 

.051 
.062 
.062 



.060 
.066 
.069 



SEMIHUMID SECTION. 



Mitchell 


1771 


Alfalfa 24 years 


Bottom 


.203 


.098 


.034 


.020 


Do. 


1772 


Cultivated 30 years 


Do. 


.180 


•093 


•033 


.017 


Do. 


1774 


Alfalfa 24 years 


Upland 


.269 


.101 


.090 


•045 


Do. 


1773 


Native pasture 


Do. 


.238 


.140 


.067 


.062 


Do. . , 


1779 


Alfalfa 23 years 


Bottom 


.160 


.068 


.062 


.048 


Do. , , 


1778 


Native pasture 


Do. 


.180 


.069 


.048 


.026 


Do. . , 


1777 


Cropped 30 years 


Do. 


.129 


.058 


.069 


.064 


Osborne 


1783 


Alfalfa 20 years 


Do. 


.184 


.101 


.062 


•039 


Do. 


1784 


Native pasture 


Do. 


.250 


.120 


.066 


.038 


Do. 


1782 


Cropped 40 years 


Do. 


.134 


.096 


.050 


•033 


Do 


1787 


Alfalfa 33 years 


Do. 


.196 


•095 


.084 




Do 


1789 


Native wood 


Do. 


.220 


•115 


.092 




Do 


1788 


Cultivated 35 years 


Do. 


•143 


.078 


.046 





SEMIARID SECTION. 



Finney 


1299 


Alfalfa 20 years 


Upland, not 


0.168 


0.085 0.047 


0.040 








irrigated 








Do 


1300 


Native range 


Not irrigated 


•137 


.084 


.038 


.048 


Do 


1303 


Alfalfa 27 years 


Bottom, irri- 


.200 


.113 


.067 


•035 








gated 








Do 


1302 


Native buffalo grass 


Not irrigated 


.135 


.086 


.050 


.031 


Do 


1301 


Cropped to grains 20 years 


Irrigated 


•134 


.101 


•058 


•033 



3IO JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Table i. — Percentage of nitrogen in Kansas soils. — Continued. 
Semiarid Section. — Continued. 



Soil 
No. 



1304 
1305 

1306 
1308 
1307 
1310 

1311 
1312 

1811 



1806 
1807 
1808 

1809 
1810 



Description of soil. 



Treatment. 



Alfalfa 27 years 
Cropped to wheat 27 years 

Alfalfa 30 years 
Native pasture 
Cropped to grains 30 years 

Alfalfa 30 years 
Native buffalo grass 
Cropped to grains 30 years 

Alfalfa 25 years 
Native range 

Alfalfa 20 years 
Native grass 
Cropped 20 years 

Alfalfa 20 years 
Native grass 



Character. 



Bottom, not 

irrigated 
Bottom, not 

irrigated 

Bottom, irri- 
gated 

Bottom, irri- 
gated 

Bottom, irri- 
gated 

Bottom, not 

irrigated 
Bottom, not 

irrigated 
Bottom, not 

irrigated 

Bottom, irri- 
gated 

Bottom, not 
irrigated 

Bottom 
Do. 
Do. 

Upland 
Do. 



Depth of sample. 



0-7 



.178 
.079 

.192 
.099 
.097 

.210 
.171 
.136 

.182 
.151 

.187 
.182 
.118 

.153 
•157 



7-20 



.080 
.057 

.082 
.052 
.066 

.085 
.108 
.079 

.101 
.088 

.107 
.115 
•057 

.066 
.072" 



20-40 



.048 
•039 

.040 
.048 
.036 

.069 
.061 
•057 

■039 
.052 



■057 
.064 
•034 

.040 
.038 



40-80 



.012 

.010 
.028 
.010 

.075 
•073 
.055 



•043 
•033 
.022 

.026 
•034 



ter. The rate of this loss decreases as time goes on. The amount 
removed by the crop depends on the size of the crop. Then there 
is a small amount of nitrogen restored to the soil by means other 
than legumes. The relative value of these as nitrogen restorers has 
been less studied than the legumes. 

In no case, except in the fields in the semiarid part of the state, has 
a continuous growth of alfalfa stored enough nitrogen to make the 
content in the soil equal to or more than that in the native sod. As 
the nitrogen content in the alfalfa soil is greater than in the soil con- 
tinuously cropped, however, it means either that the alfalfa has 
stored nitrogen so that the content is greater than it was when the 
alfalfa was sown, or that the nitrogen content has simply been main- 
tained. Which one of these is predominating must be calculated on 
certain assumptions. 

That alfalfa, like other legumes, has power to take nitrogen from 
the air is a well-recognized fact, but just how much it takes from the 
air and how much from the soil is not so well known. It is known 



SWANSON : KIKl'CT Ol" ALI-ALKA ON SOIL NITROflEN. 3II 

that alfalfa, like oIIum" Ic^^unics, does lake nitrogen from llic soil and 
that the proportion taken is i^reater in soils rich in nitrof]^en. Dr. C. 
G. Hopkins expresses the ()i)inic)n in his Soil l*\'rtilily and 1 Perma- 
nent Ac^riculture that lei^umes on the averaji^e take as much nitrogen 
from the air as is stored in the tops, and as much as is stored in the 
roots has come from the soil. This would mean that where the crop 
was entirely removed as hay, the growing of alfalfa would simply 
maintain an equilibrium. On all these fields, as far as known to the 
writer, the hay crop was removed and no manures or fertilizers of 
any kind applied. 

In Table 2 are presented the calculated results from the places 
where the three fields, alfalfa, native sod, and cultivated, were closest 
together. The analytical data, calculated as pounds per acre in the 
surface soil (0-7 inches) for the three fields is given in columns 2, 3, 
and 4. In columns 5 and 6 are given the number of years of con- 
tinuous alfalfa production and the number of years of continuous 
grain production. By subtracting the figures in column 4 from 
those in column 3, the loss of nitrogen through grain growing is 
obtained. These figures are given in column 7. The average yearly 
loss is obtained by dividing this loss by the number of years in cul- 
tivation. This average yearly loss is only approximate, because, as 
above noted, the rate is a diminishing one. This figure for the aver- 
age yearly loss is probably the best one to use for the present pur- 
pose. By multiplying the average annual rate of loss into the number 
of years the land was in cultivation before the alfalfa was seeded, the 
amount of loss sustained during the period between breaking up of 
the native sod and the seeding of alfalfa is obtained. The figures so 
obtained are given in column 9. This loss no doubt is too small, for 
the reasons given. By subtracting the figures in column 9 from 
those in column 3 the amount of nitrogen present in the soil when 
the alfalfa was seeded is obtained. The figures so obtained are 
given in column 10. By comparing the figures in column 10 with 
those in column 2, the gain or loss of nitrogen in the soil is obtained. 
These figures are given in column 11. 

Only three fields in the semiarid part of the state show a large 
gain. The losses in some of the others more than overbalance the 
gains in the rest. This simply means that the continuous growing 
of alfalfa where all the hay crop has been removed has not added 
to the stock of nitrogen in the soil. All that the growing of alfalfa 
on these fields has done, over and above grain growing, has simply 
been to prevent further losses or add enough nitrogen from the air 
to take the place of what is lost. 



312 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Table 2. — Gain or loss of nitrogen in pounds per acre in fields devoted to the 
continuous growing of alfalfa. 



County. 


Pounds of nitrogen 
per acre. 


Years 

in 
alfalfa. 


Years 

in 
grain 
crop. 


Loss 


of nitrogen- 


f^ounds 

of 
nitro- 
gen 

seeded. 


Gain 
or loss 
while in 

alfalfa 


Al- 
falfa. 


Na- 
tive. 


Crop- 
ped. 


Total 
loss by 
crop- 
ping. 


Aver- 
age 

loss. 


Loss 
before 

seed- 
ing to 
alfalfa. 


Brown 


4,220 


4.580 


3.200 


28 


45 


1,380 


31 


527 


4.050 


+ 170 


Nemaha 


3.420 


3.620 


3.200 


21 


45 


420 


9 


216 


3.400 


+ 20 


Montgomery 


2,620 


3.720 


2,200 


12 


35 


1.520 


43 


989 


2.730 


— 190 




3.360 


4,080 


2,800 


20 


35 


1,280 


37 


555 


3.530 


— 170 


Do 


3.580 


4,080 


3.260 


20 


40 


820 


20 


400 


3.680 


— 100 




3.740 


3.640 


2,360 


20 


20 


1,280 


64 





3.640 


+ 100 


Mitchell 


3.200 


3,600 


2,580 


23 


30 


1,020 


34 


238 


3.370 


— 170 




3.680 


5.000 


2,680 


20 


40 


2,320 


58 


1,160 


3.840 


— 160 


Do 


3.920 


4,400 


2,860 


33 


35 


1.540 


44 


88 


4.310 


- 390 




4,000 


2,700 


2,680 


27 


20 


20 








2,700 


+1.300 


Do 


3,840 1,980 


1.940 


30 


30 


40 








1,980 


+1,860 


Ford 


4,200 


3.420 


2,720 


30 


30 


700 


23 





3.420 


+ 780 



Further study will show that on the whole the alfalfa plant takes 
less nitrogen from the air than is stored in the leaves and stems. 
The harvesting of the crop as hay does not mean that all that which 
has grown above the soil is removed. Anyone who is at all famihar 
with the process of alfalfa hay making knows that, in spite of the 
most careful methods, large losses of leaves occur. These leaves con- 
tain on the average over twice the percentage of nitrogen that is 
present in the stems. Approximately half of the plant is leaves. In 
an experiment at the Kansas Agricultural Experiment Station, the 
loss of leaves, calculated from differences in nitrogen content in the 
alfalfa sampled as hay with usual loss of leaves and that sampled 
without loss of leaves averaged all the way from 7 to 25 percent of 
all the leaves, or from 3 to 14 percent of the entire crop. It was 
also noted that the loss of leaves was larger in a dry season than in 
a wetter one. These experimental data are in accord with prac- 
tical experience. A man who has had very extensive experience 
with growing alfalfa under irrigation stated to the writer that it was 
a most difficult crop to save and that the greatest difficulty was to 
prevent the loss of leaves. 

If alfalfa had taken as much nitrogen from the air as is stored in 
the plant above ground the soils should have shown a gain of nitro- 
gen. It is safe to assume that in practical hay making 20 percent of 
the leaves are lost. When the cutting is delayed through bad 
weather many fall to the ground before the crop is cut. Experi- 
ments have shown that in bad weather as much as half the crop is 



SWANSON : EFKF.CT OF ALFAIJ- A ON SOU. NITUOCKN. 313 

lost. The nitrogen added to the soil in these ways should produce 
an increase in the total amount present. The leaves contain 3.5 
percent or more of nitrogen. A yield of 5,cxx) pounds per annum is a 
safe estimate. If the leaves are half the crop and 20 percent is left 
on the ground this amounts to 17.5 pounds of nitrogen annually, or 
350 pounds in twenty years. In the soils sampled, except in the 
semiarid section, the calculated losses more than ofifset the gains. 

A few words of explanation in regard to apparent increased crop- 
producing power of the soil after it has been in alfalfa for some 
time is desirable. This is entirely possible without an increase in the 
stock of potential fertilizing elements of the soil. The crop-produc- 
ing power of the soil depends on several factors. Any soil which 
has been used for a perennial crop has an improved physical condi- 
tion. The alfalfa roots contain a large amount of organic matter, 
and as this decays many benefits to the soil follow. Attention may 
be called to the fertilizing elements in alfalfa roots.^ The average 
percentages of the most important fertilizing elements were as fol- 
lows : Nitrogen, 2.10; phosphorus, 0.19; potassium, 1.34; and calcium, 
0.62. The average percentage of these elements in the alfalfa plant 
as a whole, one tenth bloom, was: Nitrogen, 2.63; phosphorus, 0.18, 
potassium, 2.82; and calcium 1.07. On the whole the roots are 
poorer in these elements than the top. When an alfalfa field is 
plowed up large amounts of these elements become available to suc- 
ceeding crops. What the alfalfa has done for the soil is to prepare 
a large amount of available plant food. This is what causes the 
increased productiveness, not necessarily indicating that the poten- 
tial plant food has been increased. 

The importance of these results cannot be over-emphasized. Al- 
falfa has been and is one of the greatest crops in this and adjoining 
states. Its potential power to add not only to the sources of stock 
feed but also to the fertility of the soil is great. But these results 
show that the fields which produce the alfalfa do not get this benefit. 
Whatever nitrogen from the air is transformed into compounds 
available for grain crops is transferred to other places. If it is not 
wasted after this transference all would be well, but even super- 
ficial knowledge of present conditions leads one to believe that most 
of it is wasted. Radical changes in some practices connected with 
conserving of soil fertility are needed. 

^ These roots were dug by Prof. Ralph Kenney in cooperative work con- 
ducted at this experiment station. 



314 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Summary. 

1. Kansas has a number of alfalfa fields which have been con- 
tinuously in this crop for twenty to thirty years or more. The older 
fields are found in the central and western part of the state. Near 
these fields generally are fields which are in native sod used as pas- 
tures or as hay land, and fields which have been used continuously 
for grain growing for thirty to forty years or more. By sampling 
these fields in close proximity, data are secured from which the in- 
crease or decrease in the nitrogen content of the soil in alfalfa can 
be calculated. 

2. By assuming that the fields now in alfalfa had the same nitrogen 
content originally as the field now in native sod and that the average 
annual rate of loss before the alfalfa was seeded was the same as 
that of the fields used for continuous grain growing, the nitrogen 
content at the time the alfalfa was seeded can be calculated. By com- 
parison with the results of the three fields at the present time, calcu- 
lation can be made of the increase or decrease of nitrogen content 
due to the growing of alfalfa. 

3. In no fields in alfalfa -is the nitrogen content equal to that in 
native sod, except a few in the semiarid portion of the state, where 
it was greater. In most cases in the central and eastern parts of the 
state the nitrogen content of the alfalfa field is greater than that of 
the field used for continuous grain growing. By accounting for 
that lost before the alfalfa was seeded and comparing with the 
amount present in the soil now, it is found that on the whole the 
growing of alfalfa has not added to the amount present in the soil, 
except in a few fields in the semiarid portion of the state. All that 
the alfalfa has done has been to prevent further losses or, in other 
words, to maintain an equilibrium. 



LOVE & WENTZ : EAR C H A K ACTl^RS AND YIELD OF COKN. 315 



CORRELATIONS BETWEEN EAR CHARACTERS AND YIELD 

IN CORN.^ 

H. H. Love and J. B. Wentz. 

Introduction. 

In the early history of the corn show as an element in our agri- 
cultural education, the judges were confronted with the problem as 
to what points should be taken into consideration and the relative 
weight which should be given each point in placing one sample of 
corn above another. In 1886 the judges at the corn exposition in 
Chicago prepared a scale of points to be used at that exposition and 
from this the score card, based upon an ideal type, was developed. 
After the score card had come into general use a few experimenters 
conducted tests in which they compared the yields obtained from 
selected seed ears varying in the characters emphasized in the score 
card. In some of these tests the seed ears were selected for several 
generations for the two extremes in each character studied and the 
yields of the selected ears were compared. 

The data in this paper deal with the correlation of seed-ear char- 
acters and yield when the seed ears are not selected for extremes in 
the particular characters studied, but are nearer the average ear 
type. The purpose of the paper is to throw some light on the 
question as to whether a grower should select seed ears that have, 
for example, a certain number of rows of kernels, or a certain length 
of ear, or a cylindrical or tapering ear. 

Earlier Work. 

Williams^ of the Ohio station has conducted rather extensive and 
interesting experiments in which he selected seed ears for a number 
of years for the extremes in such characters as length of ear, shape 
of ear, filling of tip, indentation of kernel, weight of ear, and per- 
centage of grain. In selecting long and short ears he obtained a dif- 
ference of only 1.39 bushels per acre in lo-year average yield in favor 
of the long ears. In selection for shape of ear the tapering ears 

1 Paper No. 63, Department of Plant Breeding, Cornell University, Ithaca, 
N. Y. Received for publication May 21, 1917. 

2 Williams, C. G. Corn experiments. Ohio Agr. Expt. Sta. Bui. 282. 1915. 



3l6 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

excelled cylindrical ears in average yield in a 9-year period by 1.65 
bushels per acre. Eight years' selection for bare as compared to 
filled tips gave an average difference of 0.34 bushel per acre in favor 
of filled tips. Seven years' selection of rough as compared to smooth 
dented ears gave an average difference of 1.76 bushels per acre in 
favor of the smooth type. Seed ears averaging 88.16 percent of 
grain gave a 6-year average yield of 64.64 bushels per acre as com- 
pared to 65.06 bushels from ears averaging 76.38 percent of grain. 

Hartley,^ in a tabulation of yields of four varieties for six years, 
embracing 1,000 ear-to-row tests of production, found no relation 
between the characters of seed ears and yield. 

Montgomery* of the Nebraska station selected seed ears continu- 
ously for such characters as shape of ear, shape of kernel, and size 
of ear. He concludes that the results favor slightly a rather long 
seed ear, that size of ear depends too much upon environment to be 
of any importance, and that a medium depth of kernel is to be pre- 
ferred. 

Pearl and Surface^ of the Maine station conducted experiments 
with sweet corn in 1907, 1908, and 1909, making a large number of 
ear-to-row tests. Two types of commercial sweet corn were used 
and a complete study made of a number of characters. While this 
work was not primarily a study of the relation of seed-ear characters 
to yield, one of the conclusions drawn from the data obtained was 
that there is no relation between the external seed-ear characters and 
yield in sweet corn. 

Sconce^ of Illinois, experimenting with Reid Yellow Dent and 
Johnson County White corn for five years, found that seed ears of 
these varieties with 18 and 20 rows of kernels gave better yields than 
those with more than 20 or less than 18 rows. Averaging four 
years' results with these two varieties it was found that in the case 
of the Reid Yellow Dent the kernels having small germs gave better 
yields, while in Johnson County White the kernels with large germs 
yielded best. In correlating yield with shape of kernel it was found 
in both varieties that square-shouldered kernels showing a small 
space between rows at the crown and tip gave the best yields. 

3 Hartley, C. P. Progress in methods of producing higher yielding strains 
of corn. U. S. Dept. Agr. Yearbook, p. 309. 1909, 

4 Montgomery, E. G. Experiments with corn. Nebr. Agr. Expt. Sta. Bui. 
112. 1909. 

5 Pearl, Raymond, and Surface, Frank M. Experiments in breeding sweet 
corn. Maine Agr. Expt. Sta. Bui. 183. 1910. 

6 Sconce, H. J. Scientific corn breeding. Proc. Amer. Breeders' Asso., 7: 
43- 1911. 



LOVE & WKNTZ: EAR CHARACTERS AND YIELD OF CORN. 3I7 



McCall and Wheeler^ of Ohio State University calculated the 
correlations between a few ear characters and yield. The data were 
taken from some results obtained at the Ohio station and the correla- 
tion coefficients arc given in Ta1)lc i. 

Table i. — Correlations betzvcen ear characters and yield of corn, as reported 
by McCall and Wheeler. 

CoefTicients of correlation. 
Characters correlated with yield. Series i. Series 2. 

Length of seed ear 0580 + .0296 .1017 + .0651 

Weight of seed ear — .0270 + .0292 .0866 + .0656 

Circumference of seed ear — .0968 ± .0287 .1803 + .0636 

Density of seed ear 0272 + .0293 

The seed ears used in this work had not been selected for the char- 
acters used in the correlations. 

In 1 91 2, the senior author of this paper published the results of two 
years' study with two varieties.^ He correlated the characters of 
length, weight, number of rows, weight of kernel, ratio of tip cir- 
cumference to butt circumference, and percentage of grain on the 
seed ear with yield per stalk. The correlations obtained are given 
in Table 2. 

Table 2. — Correlations between ear characters and yield of stalk, as reported 

by Love. 



Seed-ear characters correlated 
with yield per stalk. 


Minnesota No. 13. 


Funk Ninety Day. 


1909. 


1910. 


1909. 


1910. 


Weight 

Weight of kernels 


— .099±.076 
.094d=.076 

.260±.072 


.241 ±.064 
.015 ±.068 

— .i27±.o67 
.028 ±.068 

— .162 ±.066 

— .i77dh.o66 


.300±.o6i 
.323 ±.060 
— .061 ±.069 


.058 ±.067 
.090±.o67 
-.034±.o67 
.043 db. 067 

.oi4±.o67 


Ratio of tip circumference to 






Percentage of erain 













The only characters that show any considerable correlations with 
yield in Table 2 are length and weight of seed ear. Of the eight 
correlations obtained for these two characters in the two years, two 
are about five times their probable errors, while others are actually 
less then their probable errors, so that these correlations can hardly 
be considered significant. In his work Love used Funk Ninety Day 

McCall, A. G;, and Wheeler, C. S. Ear characters not correlated with yield 
in corn. Jour. Amer. Soc. Agron., 5: 117. 1913. 

8 Love, H. H. The relation of certain ear characters to yield in corn. Proc. 
Amer. Breeders' Asso., 7: 29. 1912. 



31 8 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

and Minnesota No. 13 and the same methods were employed as are 
used in this paper. The present paper really is a continuation of 
this earlier work, using the Funk Ninety Day^ over a period of five 
years. In addition to the characters studied in the earlier paper 
other characters have been correlated with yield. The correlation 
coefficients were calculated in the usual manner and carefully checked 
by independent workers. 

Cunningham^^ of the Kansas station recently published a paper re- 
cording studies of the relation of seed-ear characters to yield in a 
number of varieties. The material used was from some ear-to-row tests 
that had been conducted by the station in its corn improvement work. 
The seed ears were classified for the characters studied and the 
average yield of each class determined. There was some variation 
for length of ear in the different varieties. In the small varieties 
the long ears yielded a little better than the medium and short ears, 
but this did not hold true in the larger varieties. There was no sig- 
nificant difference in the averages for all varieties. Ears with small 
circumference outyielded the larger ears. There seems to be no rela- 
tion between the filling out of the tips of the seed ears and yield. 
Smooth ears outyielded the roughly dented ears. Ears with low 
percentage of grain yielded slightly higher than those with the higher 
percentages of grain. Ears with 16 and 18 rows of kernels generally 
produced the highest yields, although there was a difference in varie- 
ties in this character. 

Material Used in the Present Work. 

The corn used in this work is Cornell No. 12, a selection from 
Funk Ninety Day, a yellow dent variety obtained from the Funk 
Bros. Seed Co. in 1908. In 1908 and 1909 it was grown on G. R. 
Schauber's farm at Ballston Lake, N. Y. In 1910 half of each 
seed ear was planted in a plat on Mr. Schauber's farm and the other 
half in a new plat on Broad Brook Farm at Bedford Hills, N. Y., 
owned by the late Seth Low. The data since 1909 came from the 
latter plat. 

Methods. 

The characters studied are length, average circumference, average 
cob circumference, and weight of ear; number of rows; average 

^ This corn has been selected for some time for earliness and yield. In part 
it is different in these two characters from the original sort and is now called 
Cornell No. 12. 

10 Cunningham, C. C. Relation of ear characters of corn to yield. Jour. 
Amer. Soc. Agron., 8: 188. 1916. 



LOVE & WENTZ : EAR C 1 F AR AC TKKS AND YIELD OF CORN. 3I9 

weight, average length, and average width of kernels ; and percentage 
of grain. All the measurements were taken in centimeters and all 
weights, excepting yield, were taken in grams. The yield of each 
row was taken in pounds and the yield per row divided by the number 
of stalks in the row to obtain the yield per stalk. The average cir- 
cumferences of the ear and of the cob were obtained by averaging the 
tip and butt circumferences. The ratio of the tip circumference to 
the butt circumference was obtained by dividing the tip circumfer- 
ence by the butt circumference. In this way the shape of the ear 
was determined. The lower this ratio, the more tapering would be 
the ear. To determine the average weight of kernels, the number of 
kernels per ear was calculated by multiplying the number of kernels 
per row by the number of rows on the ear, and then dividing the 
weight of grain by the number of kernels. The average length and 
width of kernels was determined by taking the measurements of lo 
kernels and averaging these measurements. The percentage of grain 
was calculated by dividing the weight of shelled corn by the weight 
of the ear. 

Results. 

Table 3 shows the correlations obtained for all the characters for 
each of the five years. In this table the correlations obtained by 
Love, w^orking with this same lot of corn in 1909 and 1910, are also 
included, so that correlations for seven years are shown for some of 
the characters. 

The only very significant correlations in Table 3 are seen where 
the average circumference of the seed ear is correlated with yield 
and these are not consistently high. These correlations seem to show 
some slight relation between the circumference of the seed ear and 
yield, indicating that the larger the circumference the greater the 
yield per stalk. It is noted that the correlations for average circum- 
ference of cob and for weight of seed ear are all positive, although 
they are all small. This may mean that there is a slight relation 
between the size of the seed ear and yield, so that the larger ear may 
tend to give slightly the larger yield. However, length of seed ear 
shows no constant relation to yield. 

It is interesting to note that the correlations between the percen- 
tage of grain in the seed ear and yield are all negative. While these 
correlations are all too small to be of any significance they rather con- 
firm the results Williams obtained in studying this character. It is 
not thought necessary to publish the correlation tables in detail in 
this paper. These correlations on the whole are so low that it seems 



320 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



O r^t-o t-^vooooooooo 
oooooooooo 
-H-H4^-i^4^-H-H-H-H-H 

vO rorooo ooooo ro^vO 

MMMMOlHOOOO 



oooooooo 

-H-H^^-H-H^^-H-H 

w Ooo rovo'^ioo 
OmOmmOmO 



I I I 



0\M O ^uoO\OnO\ 

oooooooo 
-H-H-H4^-H-H-H-H 

<NOO\TfOsi-i(Nro 

OsO M t^Tj-lOMOO 

MfOONtNOi-iO 



o o o o o o o o o o 

OOOOOOOOOO 

^^^^-H-H^^^^-l^-H^^-H 

Mrfoo!NOr^>-iONrovo 
0(NOihmmihOOm 

f ' i' " ' r r ' * ' 



oooooooo 
-H4]-H-H-H-H-H-H 

MoorO^iNOOM 
OmOmOmmm 

' ' \ ' ' \ \ ' 



111 



03 o ^ ' 



O O 



^ < 



^ 2:1 'S 



, ^ o 

a; >-i 

- s 

>l 



OJ (LI 
> > 
< < 



■^(NOi-Ht— r^OOOOO 
00000\000000 



-H-H-H4^4^-H414i4^-H4i 

<N00 0\ O\0O 

00 t^uooo •^tooo O M C><N 
(r)<Noo<NO<N(NOroi>i> 



0\iO0 rfoo ^O lOO O O 
OOOOiNmOhOOO 



-H-H-H-H-H4^4^-H4^-H4^ 



^ O (N mo -rf 



MOOrO'^OOfO 
MOOO-<:tMOM 



-H-H-+1-H-H-H-H-H 

OOOi^'OOmO 
(N TflOlOO -rf M <N 
OsOOO (MOO q\<T)QO 

M uo ' d\ 00 u-> ' o 

(N M 0\ 00 M 



Ov'OO'^^vOO-^ 
OOOO^mOm 



OOTl-OOt^oOO 
rO^iJOOO O -^OO-O 
00 rooo M Tt-0\<NO 



0\iO0 ro^r^O M o o o 
OOOOOOhOmOOO 



-H-n^^4^4^4l-H-H-H-H4^ 

OOOOOOloO'*(NC^ 

roM r^O OO 0\(N m 

(N "^00 (NOOO <N rot^O 



OONroiOOONro 
0\U-)0 "^t^-ioO fO 
OOOOOmOw 



^^4^-H4^-H-H-H-H 

OOmOOOOO 
m(NO\<NOO<N00 
rOiOOOO'^MC^JOv 



o 
u 



ill 



o 

^ O 



OJ > > 



.2r 

> Ah 



LOVE & WENTZ: EAR CHARACTERS AND YIELD OF CORN. 32 I 

there is no relation between the seed-ear character and yield ; at 
least, there is not enough evidence to justify one in attempting to 
improve the yield of his corn by selecting any particular type for seed. 

In Table 4 the mean percentages of grain for the ears in a few of 
the highest classes and a few of the lowest classes in the correlation 
tables are shown, with the mean yields of these classes. In this table 
it will be noticed that in every case the seed ears with the lowest per- 
centage of grain gave slightly higher yields. 



Table 4. — Mean percentages of grain in the seed ears in a few of the highest 
and lowest classes, with the mean yields of these classes. 





High classes. 


Low classes. 


Year. 


Percentage of 


Yield per stalk, 


Percentage of 


Yield per stalk, 




grain. 


pounds. 


grain. 


pounds. 




87.074 


0.822 


81.676 


0.859 




88.565 


0.624 


79.375 


0.677 




87.158 


0.672 


80.750 


0.738 


I913 


88.447 


0.692 


80.714 


0.847 


I914 


87-235 


0.651 


82.750 


0.645 


Average 


87.596 


0.692 


81.053 


0.753 



In addition to the correlations between the seed ear characters and 
yields, the means, standard deviations, and coefficients of variability 
were calculated for the seed ears for each year. It may be of in- 
terest to note the means of the characters studied. These constants 
are shown in Table 5. 

In Table 5 there seems to be a slight tendency for the ears to grow 
longer. The ratio of ^ip circumference to butt circumference has 
decreased slightly, meaning that the ears seem to have become more 
tapering. The weight of the ears has increased considerably. In 
looking at the means of the yields it may appear as though there has 
been a slight decrease in yield, since the yield in 1910 was so much 
higher than in any of the following years. This higher yield can, in 
most part, be accounted for by the fact that in this year the corn 
was not so mature as it has been in the succeeding years. Consider- 
able green corn was produced the first year, while the following year 
at harvest time the corn was nearly all mature. 

Conclusions. 

I. The characters of length, ratio of tip circumference to butt cir- 
cumference, average circumference of cob, weight, average weight of 
kernels, number of rows of kernels, and average length and width of 



322 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

kernels on the seed ears do not show correlations significant enough 
to be of value in judging seed corn. 

2. The data indicate a slight negative correlation between per- 
centage of grain in the seed ear and yield, meaning that possibly ears 
containing a low percentage of grain yield higher than ears with a 
high percentage of grain. 

3. The average circumference of the seed ear is the only character 
that shows any significant relation to yield. 

4. The judge at a corn show or a farmer in selecting his seed corn 
cannot pick the high-yielding seed ears when judging from outward 
characters of the ears. It is evident that the points emphasized on a 
score card are of no value for seed ear purposes and are entirely for 
show purposes. 

5. The only basis left for selecting high yielding seed corn is the 
ear-to-row progeny test. 



VEGETATION ON SWAMPS AND MARSHES AS AN INDI- 
CATOR OF THE QUALITY OF PEAT SOIL FOR 
CULTIVATION.! 

T. J. DUNNEWALD. 

The general conclusion of various observers has been that the 
surface vegetation on peat gives no clue to the relative quality of 
the soil for purposes of cultivation. Among the reasons given as 
to why this is so, the following are important: 

That in the natural development of a peat bog there is a more or 
less rigid succession of plant types and that tamarack and spruce 
timber indicate only that the deposit is in a mature stage of its growth. 

That such accidents as fire, epidemics of plant disease or pests, 
changes from a series of dry years to one of wet years, changes in 
the elevation of the water table of the deposit, etc., may serve to 
destroy entirely one type of vegetation and create conditions more 
favorable to very different kinds of plants. 

On the other side of the question are the oft-repeated assertions 
of farmers and drainage men that " the peat on a black spruce or 
moss-covered swamp is no good for cropping and money spent in 
draining it is wasted," while "good black muck with elm or ash on 
it is the best kind of land to drain." 

1 Contribution from the Wisconsin Agricultural Experiment Station, Madi- 
son, Wis. Received for publication March 2, 1917. 



dunnewald: swamp and marsh vi:(;ktation. 323 

One of the duties of the writer has l)een to examine, map, and 
write a preliminary report on the soil of proposed drainage distriets 
under the State drainaj^e law. One such area examined consisted 
of a township in the northwestern county of Wisconsin, which con- 
tained ahout 6,600 acres of scattered marshes and swamps. 

The upland is non-calcareous glacial drift derived from granitic 
and sandstone rocks, with no limestone in the vicinity. Admitting 
that there might be mechanical and physical conditions in the peat 
soil of these areas which might make it favorable or unfavorable 
for cultivated plant growth, the writer wished to know whether 
a favorable or unfavorable chemical condition could be found in dif- 
ferent areas of the peat and whether the vegetation would parallel 
any such condition. 

In addition to a careful field examination, representative samples 
of the peat bearing different classes of vegetation were taken to the 
laboratory for examination. 

As a result of the field study alone it was concluded that the 
spruce and tamarack peat areas were the wettest, with the water- 
table practically at the surface of the soil and a covering of 12 to 
18 inches of spongy moss. The depth of the peat or distance from 
shore seemed to affect the kind of surface growth but little, and the 
extent of the decomposition of the peat, that is, its fine grained or 
fibrous condition, had but little more effect. The rawest samples 
of peat were found on the spruce and tamarack areas, but as often 
the peat was as well decomposed under spruce and tamarack as under 
elm, birch, ash, and grass. 

The examination of the samples in the laboratory gave somewhat 
more positive results. Table i gives a summary of the more inter- 
esting determinations which were made. These data indicate that 
the peat bearing black spruce and tamarack has 20 percent less min- 
eral matter, a much greater degree of acidity, and somewhat less 
nitrogen. 

While the greater acidity present in the spruce and tamarack peats 
may be due to more continued flooded conditions on those swamps 
and while drainage experience shows that this acidity often disappears 
largely after the drainage and cultivation of the peat, we believe the 
data support the farmer's statement that such trees as ash, elm, birch 
and white pine on peat indicate a better quality of the material than 
that where only black spruce, tamarack, sphagnum moss, blueberries, 
and Cassandra grow. 

Determinations of the solubility of the peats in 150 cc. of 2 per- 



324 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



cent NaOH solution showed that the acid peats were from 3 to 8 
percent more soluble than the less acid ones, but when sufficient 
excess amount of the solvent to counteract the extra acidity of the 
spruce peats was used, the difference in solubility was not noticeable. 



Table i. — Comparative determinations on peat soils in the same locality bearing 
different classes of vegetation.<^ 



No. of 








Total 


Vegetation, 


Loss on 


Truog acidity. 


nitro- 


soil.* 


ignition. 


gen. 






/O 






2 






\/prv Qfrnno' 

V y OLJL^Xl^ 


I ^8 




T3,m3,r3,ck and moss 


TO 1 1 
/y. J. i 


Strong 


1.98 






78.80 


Strong 


I '78 


5 


Black spruce and moss 


85.48 


Very strong 


1.86 


8 


do 


91.07 


Very strong 


1.86 


9 


do 


90.89 


Very strong 


1.90 


II 


do 


88.14 


Very strong 


1.69 


14 


do 


93.01 


Very strong 


1.82 




Average 


88.90 


Very strong 


1. 81 


I 


Mixed ash, birch and balsam 


60.61 


Very slight 


1.96 


3 


Large ash, birch, poplar, and 










cedar 


81.01 


Slight 


2.17 


6 


Birch, ash, elm 


66.91 


Medium 


2.02 


7 


Mixed birch, ash, tamarack, and 










willow 


'^56.85 


Slight 




10 


Ash, birch, and a few large 










tamarack and pine 


47.14 


Medium 




17 


Elm, ash, cedar, and grass 


62.08 


Medium 


2.20 




Average 


67.60 


Slight 


2.09 



o The peat was from 6 to 20 feet deep wherever a sample was taken except in 
the case of No. 7. 

& Samples Nos. 2 and i were taken in different parts of the same swamp, 
c Fine sand grains were detected in Nos. 7 and 10 and they are not included 
in the average loss on ignition. 



GERICKK: KI'KECTS of cropping to llAKl.KV. 



325 



SOME EFFECTS OF SUCCESSIVE CROPPING TO BARLEY.^ 

W. F. Gericke. 
Introduction. 

The work reported in this paper is part of an experiment which had 
as its purpose the study of some of the effects of successive cropping 
of barley on a soil in pots under greenhouse conditions. All the 
factors of production except the soil were the same or similar for 
the series throughout the experiment. The experiment was carried 
out as follows : Portions of 5 kilograms of soil were weighed out 
and placed in pots. Some of the pots were then sown while others 
remained unsown until a series was obtained some of which had 
produced three crops, some two crops, some one crop, and some no 
crop. Throughout the experiment both the culture and water con- 
tent of the soil for the series were kept similar whether the pots grew 
plants or not. 

The soil used in the experiment was a Berkeley hillside adobe to 
which some compost material had been added, and was in good phys- 
ical condition. Proper care was taken to insure uniformity in type, 
texture, and composition. A sample of this soil upon analysis by the 
strong acid digestion method gave the following results : 



Insoluble residue 64.85 

Soluble silica 9.18 

CaO 2.26 

Fe.Os 4.59 

ALO3 5 80 

SO3 04 

MngO* 13 

MgO 72 

K,0 62 

Na,0 43 

48 

Loss on ignition ii-94 

Total 101.04 

Total nitrogen 31 

Humus 3.20 

Nitrogen in humus 3.30 



^ Contribution from the Laboratory of Soil Chemistry, University of Cali- 
fornia, Berkeley, Cal. Received for publication May 4, 1917. 



326 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

The analysis shows that the soil was well supplied with the neces- 
sary plant-food constituents and that the nitrogen content was es- 
pecially high. All of these factors must be kept in mind in reading 
the paper. 

The seed used in the experiment was a very pure strain of Beldi 
barley. All the seeds planted in the pots came from two plants. 
After the seedlings were about 3 inches high all pots were thinned to 
two plants each, one from a seed from each of the plants used. 

Experimental Results. 

The data recorded in the following tables includes length of period 
of harvest, tillering, height of stalk, and weight of grain for the in- 
dividual heads, and average weight of kernel per head. The tables 
are arranged to bring out certain relationships. They are (a) those 
showing the different kinds of stalk production and maturation of 
the crops; (b) those showing the total and average height of the 
different kinds of stalks of the crops; and (c) those showing the 
quantity and quality of grain production as related to the height of ^ 
the stalks in each of the crops. 

STALK PRODUCTION AND MATURITY. 

The data on total number of stalks, stalks producing grain, and 
stalks producing heads but no grain, with the dates of first and last 
ripening of the various crops, are shown in Table i. 



Table i. — Stalk production and maturation of barley as affected by successive 
cropping of the same soil. 





Number of stalks. 


Period of harvest. 


No. of 






Producing 
heads but no 








crop. 


Total per 


Grain pro- 


Producing no 


First heads 


Last heads 




pot. 


ducing. 


grain. 


heads. 


ripe. 


ripe. 


4 


5 


5 








June I 


June II 


4 


6 


6 








do. 


do. 


3 


20 


6 


6 


8 


June 15 


July 7 


3 


12 


6 


4 


2 


do. 


July 8 


2 


21 


6 


10 


5 


June 15 


July 10 


2 


23 


8 


9 


6 


June 14 


do. 


I 


29 


13 


II 


5 


June 15 


July 8 


I 


40 


10 


16 


14 


June 16 


July 12 



Briefly discussing the data presented, Table i shows a difference 
in the length of the period of harvest, i. e., in the maturation of the 
heads. The plants in crop 4 had the shortest growing period because 
the period of maturation was much less for them than for those in 



gericke: eeeects ok cropping to barley. 327 

any other pots. Due to good uniformity in the ripening of the 
grain in crop 4, all heads produced were harvested about the same 
time. This fact, however, must not convey the impression that all 
heads had matured to the same degree of ripeness, a fact which 
could only be ascertained by chemical analysis. The practical cri- 
terion of ripeness in the color and hardness 'of grain was such as to 
indicate the advisability of harvest. In respect to the uniformity of 
maturation of the grain, the plants in crop 4 must be considered as 
more desirable than any of the other plants. The importance of 
uniformity and even ripening of grain in the field is of the utmost 
importance. 

The plants representing the first, second, and third croppings of a 
soil had a very extended period of harvest. Although some heads 
matured a few days after those of crop 4, the fact that a period of 
three to four weeks was required for all heads to ripen presented 
conditions which if duplicated in the field would be very serious to 
farming practices. It is quite true that these conditions of pro- 
longed maturation of grain could hardly be expected under field con- 
ditions when a pure strain of seed is used. Still, the fact remains that 
the soil conditions are important factors in contributing to uni- 
formity and evenness in the ripening of grain. A glance at Table i 
shows that the plants which required a relatively long period of time 
for maturation were plants that produced a considerable number of 
barren stalks. The excessive tillering of plants, noted especially in 
the first crop, but also evident in those of the second and third crops, 
undoubtedly are related to soil conditions and these factors have 
brought about the extended period of harvest. A partial explanation 
of the long harvest period of these series, however, may be found 
in the effect of the removal of ripe heads from the plants, thereby 
allowing a fuller development of later-formed heads on the tillers. 

While a moderate amount of tillering is a very desirable feature 
in grain production, too much vegetative growth in cereals often re- 
sults to the detriment of grain production. As a tiller will produce 
grain if conditions are favorable, the time during the plant's growth 
when tillers appear is of great importance in procuring uniformity 
and evenness in the ripening of grain. 

From the standpoint of the most efficient tillering of barley for 
grain production, in which the quality and quantity of grain are the 
factors of prime importance, the produce of the fourth crop must be 
considered the best, as a glance at Table 3, which gives the total 
weight and average weight of the kernels, shows. Judging then 



328 . JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

from the standpoint of the fertility of a soil for certain features of 
crop production, so far as this experiment is concerned, the soil at 
its fourth cropping was more fertile for grain production than at 
its first. 

The maximum dry matter production was attained in the first crop. 
The straw production in this crop was more than lOO percent larger 
than that in the fourth crop, but the grain yield exceeded that of the 
fourth crop by only a small margin. In the ratio of percentage yield 
of grain to total dry matter production the plants from the fourth 
cropping were far superior to those from any other crop. In the 
quality of grain as indicated by the average weight per kernel, the 
production of the fourth crop exceeded that of the first by a margin 
of 6^ mg. per kernel, which would make a very significant differ- 
ence in the grading of the grain produced from the several crops. 

Undoubtedly soil conditions were responsible for the differences 
noted in the plants of the several croppings. With the plant food 
available in different amounts and ratios of one element to another 
in the various crops, a condition was brought about that was re- 
flected in the quantity and quality of the produce. In the first crop 
the supply of plant food, both in quantity and in the ratio of avail- 
able elements one to another, was of such a magnitude as to induce 
an extended vegetative growth. Thus tillering was not only induced 
but sustained to fruition, over a relatively longer period of time. 

HEIGHT OF STALKS. 

Table 2 shows the total and average heights of the stalks of the 
different classes. 



Table 2. — Height in centimeters of the different classes of stalks as affected by 
successive cropping of the same soil. 









Height of grain- 


Height 


of stalks 


Height 


of stalks 


No. of 


Height of all .stalks. 


producing stalks. 


producing heads, 
but no grain. 


producing no heads. 


crop. 


















Total. 


Average. 


Total. 


Average. 


Total. 


Average. 


Total. 


Average. 


4 


343 


68.6 


343 


68.8 










4 


400 


66.6 


400 


66.6 










3 


815 


40.7 


319 


53.2 


175 


29.2 


321 


40.1 


3 


471 


38.2 


255 


42. 5 


110 


27-5 


106 


53-0 


2 


738 


35-1 


298 


49-7 


318 


31.8 


122 


24.4 


2 


732 


31.6 


330 


41.2 


266 


29.6 


136 


22.7 


I 


903 


3I-I 


502 


38.6 


284 


25.8 


117 


23-4 


1 


1. 156 


28.9 


409 


49.0 


464 


29.0 


284 


20.3 



gericke: effects of croiting to uaki.ey. 329 

The data in Table 2 show some of the effects of successive crop- 
ping of a soil on the height of the barley. The total height of all 
the stalks produced decreased for each succeeding crop, while the 
average height of the individual stalk increased. 'J'hc tallest stalks 
were found in the fourth crop plants and the shortest in the first 
crop. The average height of the grain-producing stalks was uni- 
formly greater than those of the non-producing stalks. In the first 
and second crops the barren stalks were decidedly shorter than the 
grain-producing stalks, while in the third crop the barren stalks at- 
tained their greatest height. 

Why the plants of the fourth crop produced taller stalks than those 
grown on a soil that supported plants for a less number of seasons 
is not easy to explain. That the food requirement of plants at dif- 
ferent periods of growth varies both in quantity and ratio of one 
element to another has been indicated by various investigations. The 
degree of absorptivity and availability of the elements by the plants 
in the several crops may be of importance to account for certain 
changes in the form and features of the plant. 

WEIGHT OF GRAIN. 

Table 3 shows some of the relations between the height of the 
stalks and their grain production. 



Table 3. — Relation of height of stalk to weight of grain as affected by suc- 
cessive cropping of the same soil. 

FOURTH CROP. 





Pot I. 


Pot 2. 


Stalk No. 


Height 
of stalk. 


Number 
of kernels 


Weight of kernels 
in head. 


Height 
of stalk. 


Number 
of kernels 


Weight of kernels 
in head. 




in head. 


Total. 


Average. 


in head. 


Total. 


Average. 


I 

2 

3 
4 
5 
6 


Cm. 

83 
79 
78 
62 
42 


39 
30 
32 
24 
6 


Grams. 
1.999 
1.540 
1.600 
1. 187 
.230 


Mgs. 
51-3 
51-3 
50.0 
49.4 
38.3 


Cm. 

84 
81 
73 
61 
56 
45 


37 
37 
35 
24 
26 
16 


Grams. 

1-995 
1.872 

1-703 
I. Ill 
1. 021 
.619 


Mgs. 
53-9 
50.6 
48.6 
46.3 
39-3 
38.7 


Totals. . . 
Average . 


343 
68.6 


131 
26.2 


6.556 
1.3H 


240.3 
50.0 


400 
66.6 


175 
29.1 


8.321 
1-387 


277-4 
47-5 



330 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Table 3. — Relation of height of stalk to weight of grain as affected by suc- 
cessive cropping of the same soil. — Continued. 



THIRD CROP. 



Stalk No. 


Pot I. 


Pot 2. 


Height 
of stalk. 


Number 
of kernels 
in head. 


Weight of kernels 
in head. 


Height 
of stalk. 


Number 
of kernels 
in head. 


Weight of kernels 
in head. 


Total. 


Average. 


Total. 


Average, 


I 

2 

3 
4 
5 
6 


Cm. 

74 
63 
55 
53 
48 
26 


37 
26 
22 
21 
21 
3 


Grams. 
1.921 
1.222 
1.027 
.981 
•893 
. .068 


Mgs. 

52.9 
47.0 
46.7 
46.7 
42.4 
32.7 


Cm. 

69 
64 
36 
34 
27 

25 


37 
29 
18 
21 
10 
3 


Grams. 
1.940 
1.465 
.608 
.896 
.383 
.073 


Mgs. 
52.4 
50.5 
33.8 
42.6 
38.3 
24.3 


Total 

Average . 


319 
53.1 


130 
21.6 


6. 112 
1. 019 


268.4 
47-0 


255 
42.5 


118 
19.7 


5-365 
.894 


241.9 
45-4 



SECOND CROP. 



I 


69 


50 


2.017 


40.3 


56 


33 


1.408 


42.6 


2 


56 


36 


1.080 


30.0 


53 


33 


1.390 


42.1 


3 


53 


36 


1.377 


38.2 


46 


9 


•433 


48.1 


4 


52 


27 


1.228 


45.5 


43 


17 


.742 


43.6 


5 


40 


20 


.614 


30.7 


38 


17 


.540 


31.8 


6 


28 


33 


.129 


43.0 


36 


24 


1. 000 


41.6 


7 










33 


4 


.140 


35-0 


8 










25 


I 


•034 


34.0 


Total 


298 


172 


6.445 


227.7 


330 


138 


5.687 


318.8 


Average . 


49-7 


28.7 


1.074 


37-5 


41.2 


17.2 


.711 


41.2 


FIRST CROP. 


I 


50 


42 


1.720 


40.9 


58 


34 


1.529 


44.8 


2 


50 


15 


.426 


28.4 


56 


33 


1.529 


46.4 


3 


41 


18 


•742 


41.2 


50 


22 


.977 


44-3 


4 


41 


14 


.567 


40.5 


45 


23 


1.085 


47.2 


5 


40 


20 


.804 


40.2 


41 


20 


.864 


43.0 


6 


39 


12 


.501 


41.8 


36 


10 


.409 


40.9 


7 


39 


22 


.836 


38.0 


34 


17 


•695 


40.9 


8 


36 


17 


.770 


45.3 


33 


4 


.146 


36.5 


9 


36 


7 


.299 


■ 42.7 


28 


I 


.043 


43.0 


10 


36 


8 


.309 


38.6 


28 


I 


.038 


38.0 


II 


34 


27 


1. 001 


37.0 










12 


30 


4 


.154 


38.5 










13 


30 


I 


.023 


23.0 










Total 


502 


207 


8.652 


496.1 


409 


172 


7.315 


425.0 


Average . 


38.6 


15-9 


.665 


41.8 


40.9 


17.2 


.731 


42.5 



In the plants of the fourth crop, the weight of the grain per head 
and the average weight of the kernels varied with the height of the 



gericke: eefects of ckoi'im nc, to hakf.ey. 331 

stalk, the tallest stalks producing- the most grain both in total weight 
and in the average weight of the kernels. The number of kernels 
per head, with a few exceptions, varied with the height of the stalk, 
the largest number of kernels being oji the tallest stalks. In the 
plants of the third crop, a similar relation was found between weight 
of grain and height of stalk and also between number of kernels 
per head and height of stalk. The average height of the stalks and 
the average weight of their grain was less than those of the fourth 
crop. Grain production of this crop was the least for the series, as 
these plants, unlike those of the fourth crop, produced barren stalks. 
In the first and second crop there was a decided change in the rela- 
tion of the height of stalk to the weight of grain per head and its 
average weight per kernel. While some of. the heaviest heads were 
found on the tallest stalks, many of the relatively smaller heads were 
also found on some of the tallest stalks. The heads with the best 
average weight of kernels were produced on stalks of medium height. 
The average height of the stalks and the average weight of the 
kernels were less than those of the third and fourth crops. The 
first and fourth crops had the fewest undersized kernels and also 
showed the least variation from the average mean weight of kernels 
for the crop. 

Similar to the causes given for the results in Tables i and 2, the 
differences in plant features shown in Table 3 must be ascribed to 
the soil conditions. The uniformly better grains of the fourth crop 
can only be explained 'by the fact that all stalks had progressed uni- 
formly and were within a relatively narrow range of variation for 
the respective period. Due to the absence of new stalk development 
in a later period of growth, the condition was obviated by which a 
new stalk would detract very much from the nourishment of the 
other parts of the plant, or the older stalk detract from that of the 
younger shoots. However, as the heaviest grain, both in head and 
average weight per kernel, varied with the height of the stalks in 
the fourth crop, there were obviously some soil factors at work to 
produce such a condition. While the maturity of the heads in this 
crop was such that they were harvested about the same time, the 
actual growing period of the heads of the smaller stalks was less than 
that of the taller stalks, as the heads on the smaller stalks appeared 
from one to two days later. This difference gave the taller stalks 
such a lead as to enable them to maintain their supply of plant food 
over the younger stalks, all of which were within the corresponding 
period of development. When part of the plant was in one period 



332 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

of growth and part in another, for example, when some stalks were 
headed out while others were still producing tillers, the plant food 
supplied by the same root system was influenced by a difference in 
the plant-food requirements gf the different periods. As the amount 
of water in the tissues varies with different stages of growth, the 
translocation and assimilation of nutrients is likewise affected. It 
is thus when the roots must supply food and water for two dif- 
ferent periods of development where different food requirements 
and assimilative powers exist, that no correlation is attained between 
the straw and the grain production both as to weight of grain per 
head and as to weight of the individual kernels. That the length of 
the maturation period may also effect a difference in the chemical 
composition of the grain has been indicated by other investigations. 
Soil conditions, therefore, must be considered factors that may affect 
crop production not only in quantity but also in quality and in com- 
position of its produce. 

Summary. 

1. In an experiment to determine some of the effects of continuous 
cropping of a soil under greenhouse conditions, barley was grown in 
pots, the successive crops being grown concurrently in order to 
eliminate as much as possible such factors of differences as climate 
and season. 

2. Plants of the fourth crop matured with greater uniformity than 
those of any of the other crops. There were no barren stalks in the 
plants of the fourth crop. The number of tillers and barren stalks 
increased with the plants grown in the soil of a lesser number of 
crops. 

3. The total height of all the stalks produced decreased with each 
successive crop, but the average height of the individual stalks in- 
creased with each successive crop. 

4. In the fourth and the third crops the heaviest grain, both as to 
weight per head and as to average weight per kernel, varied with the 
height of the stalks. The tallest stalks produced the largest heads 
and the largest average weight per kernel. 

5. In the second and first crops no correlation between the height 
of stalks and weight of grain per head, or average weight per kernel, 
was obtained. 



Journal of the American Society of Agronomy. 



Plate 8. 




The cells were set up in the manner here shown. The graduated pipettes 
were held in position by universal clamps, thus permitting them to be lowered 
as the solutions rose in them. The influence of a hydrostatic head was thus 
overcome. 



SCHUSTER: A STUDY OV Sf)Ii: SOLUTIONS. 



333 



i 

A STUDY OF SOIL SOLUTIONS BY MEANS OF A SEMIPER- 
MEABLE MEMBRANE SUPPORTED ON A POROUS 
CLAY PLATE.i 

Geo. L. Schuster. 
Introduction. 

The object of the work reported in this paper was to find out if 
possible the strength of various soil solutions in terms of a given 
sugar solution. Pulling and Livingston,^ in their work concerning 
the water-supplying power of the soil, used a collodion membrane 
formed by the evaporation of the solvent from a solution of " Scher- 
ings celloidin " dissolved in a mixture composed of equal parts by 
volume of ether and alcohol. Morse^ prepared semipermeable mem- 
branes of copper ferrocyanide in the porous walls of clay cups by 
using an electric current to drive the membrane-forming solutions, 
copper sulfate and potassium ferrocyanide, into the walls. 

The membranes used in this work were prepared 'by Professor A. 
G. McCall, now of the Maryland Agricultural Experiment Station, 
in the following manner. Three conical-shaped cells, having a base 
7.5 cm. in diameter and a neck 2.3 cm. in diameter and 3.7 cm. long, 
were used. The exterior surface of these cells was glazed with the 
exception of the bottom. They were placed in a dilute solution of 
Hthium sulfate (0.5 gram per liter) and a similar solution added to 
the inside of the cells. An electric current of 7 to 8 amperes was 
passed from a platinum electrode situated in the solution outside the 
cell to another inside the cell for one hour twice a day for a period of 
three days. This drove the air out of the porous bottom, but in the 
process the solution in the cell continually bubbled over, so a funnel 
was provided that the solution might be readily replaced. The form 
of the cell and the general arrangement of the electrode, etc., are 
shown in figure 21. 

After rinsing the lithium sulfate out, the cells were partially im- 

1 Contribution from Ohio State University, Columbus, Ohio. Received for 
publication April 4, 1917. 

2 Pulling, H. E., and Livingston, B. E. The water-supplying power of the 
soil as indicated by osmometers. Carnegie Institution of Washington, Pub. 204, 
P- 55-56. 1915. 

3 Morse, H. N. The osmotic pressure of aqueous solutions. Carnegie Insti- 
tution of Washington, Pub. 198, p. 82-85. 1914. 



334 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



mersed in tenth-normal copper sulfate and partially filled with tenth- 
normal potassium ferrocyanide. The cells were closed at the top 
with rubber stoppers carrying a wire attached to a platinum cathode 
(P), a funnel (S), and an overflow tube (0) (figure 21). A copper 
anode was placed in the copper sulfate outside the cell and an electric 
current of 7 to 8 amperes was allowed to pass through each cell for 
one hour each day for a week. During the periods when the current 
was on the potassium ferrocyanide solution was renewed every ten 
minutes. When the current was not on the cells were kept in dis- 
tilled water and to this a small crystal of thymol was added to pre- 
vent fungous growth. 




Fig. 21. Arrangement of electrode and solutions used in the experiment: P, 
platinum electrodes; L, lithium sulfate; S, funnel for adding new solution; O, 
overflow. Arrows at the base of the cell indicate direction of the electric 
current. 



One of the cells was broken in order to determine the location of 
copper ferrocyanide membrane in the plate. The position was found 
to vary somewhat, but the general location was in the interior of the 
clay plate midway between the two surfaces. 

Experimental. 

After the membranes had been in distilled water for five months 
their permeability was tested with molasses. It was found that the 
membranes did not hold the molasses entirely and that there was 
some exosmosis or movement of the molasses within the cell to the 
water outside the cell. This was thought to be due to a possible leak 
in the membranes, so the cells were again placed in copper sulfate 



SCHUSTER: A STUDY OV SUIL SOLUTIONS. 



335 



solution and potassium fcrrocyanidc added to ihc inside and allowed 
to remain for one week without the application of the electric cur- 
rent. At the end of that time the membranes were again tested and 
no exosmosis was detected. 

The next step was to determine the behavior of some sugar solu- 
tions of known strength. In order to avoid the influence of a hydro- 
static head, a scheme was devised by means of which the membranes 
were held in a vertical position and the surface of the liquid in the 



Fig. 22. Sketch showing the location of the mem- 
brane and connection of the cell to the graduated pipette 
by rubber tubing. S, sugar solution ; D, distilled water ; 
T, thistle tube, for supplying distilled water; Th, ther- 
mometer; P, paraffine seal; C, cork; M, 




tube kept level with the top of the distilled water in the outer con- 
tainer. The cells were supported in loX 15 cm. glass cylinders by 
means of a cork 4 cm. thick. This cork, which also served as a seal, 
was first boiled in paraffine and after insertion sealed over tight with 
the same material. Thistle tubes were provided in order that more 
water could be provided as it was taken up by osmosis through the 
semipermeable membranes of the cells. In this way the water was 
supplied to all parts of the membrane alike. 

In figure 22, S is the sugar solution in the cell, the cell being con- 
nected to the graduated pipette by means of rubber tubing ; M is the 
semipermeable membrane of copper f errocyanide ; D is the distilled 
water which is supplied through thistle tube T; C is the cork which 



33^ JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

holds the cell in place; the cork is sealed over with paraffine, P; and 
Th is the thermometer. The method of setting up the cells is 
shown in Plate 8. The graduated pipettes are held in position by 
universal clamps, thus permitting the pipettes to be lowered as the 
solutions rise in them. The influence of a hydrostatic head is over- 
come in this way, for the water in the thistle tube and the solution 
in the pipettes can be kept on the same level. 

The rates of rise for a 5, 10, and 15 percent sugar solution were 
obtained with the cells set up in the manner described above. Each 
cell was numbered and retained the same number throughout the 
series of experiments. Table i gives the rates of rise of the 5, 10, 
and 15 percent solutions, the temperature at the time of each read- 
ing, and the hours that had elapsed since the time of the first reading. 

Table i. — Rise of 5, 10, and 15 percent sugar solutions in cells with a semi- 
permeable membrane. * 



RISE OF 5 PERCENT SOLUTION IN 95 HOURS. 





Cell No. I. 


Cell No. 2. 


Time in 
hours. 




Readings on 






Readings on 






Temperature. 


graduated 


Rise. 


Temperature. 


graduated 


Rise. 






pipette. 






pipette. 






° c. 


CC. 


mm. 


° c. 


CC. 


mm. 


O 


21 








20 








20 


24 


•45 


10 


24 


1.55 


35-5 


25 


24 


.50 


"•5 


24 


2.20 


50.5 


45 


18 


.75 


17.0 


18 


4.70 


108.5 


67 


14 


1.05 


24.0 


14 


6.20 


141. 


95 


22 


1. 15 


26.5 


Out top of 10 CC. pipette. 


RISE OF 10 PERCENT SOLUTION IN 165 HOURS. 





27 








27 








16 


19 


.20 


4.0 


19 


.50 


II-5 


21 


22 


.45 


10. 


22 


.90 


20.5 


38 


22 


•95 


22.0 


21 


1.70 


39-0 


64 


10 


I-I5 


26.5 


II 


2.20 


50.5 


92 


21 


1-95 


46.0 


21 


3-40 


78.5 


118 


20 


2.45 


57^0 


20 


4-25 


98.0 


142 


17 


2.80 


64-5 


18 


4.90 


113. 


165 


21 


3-35 


77.0 


22 


5-90 


136.0 


RISE OF 15 PERCENT SOLUTION IN I70 HOURS. 





25 








25 








24 


23 


•65 


15.0 


21 


.65 


15-0 


32 


24 


•95 


22.0 


24 


•95 


22.0 


48 


22 


1-45 


33-0 


21 


1.40 


32.0 


58 


20 


1.70 


39-0 


20 


1.70 


39-0 


75 


22 


2.30 


53-0 


21 


2.35 


54-0 


82 


23 


2.60 


60.0 


22 


2.70 


61.0 


96 


20 


2.90 


69.0 


19 


3.10 


72.0 


121 


21 


3-70 


88.5 


20 


4.10 


86.0 


145 


17 


4-25 


99.0 


17 


4.70 


110. 


170 


19 


5.00 


117. 


20 


5-05 


118. 



sciiustkr: a siri)\- oi- son. solutions. 



337 



After the testing- of the iiKMiibranes with each of the 5, 10, and 15 
percent sug^ar sohitions, the cells were again placed in a solution of 
copper sulfate and a solution of potassium ferrocyanide added to the 
inside. They were allowed to remain in contact with the solutions 
for a period of one week before the next test was made. This was 
done to repair any leaks that might have developed in the membranes. 

It will be noted from an inspection of Table i that the stronger the 
solution the more steady and uniform the rise; also, that the two 
membranes behave more nearly alike in the stronger solutions. This 
may be due in part to the ageing of the membranes and in part to the 
additional soaking and their consequent renewal in copper sulfate and 
potassium ferrocyanide solutions. 

Operation of the Membranes Against the Soil. 

The types of soil used in this work were sandy loam, muck, and 
clay. The samples were taken from the surface 6 inches and put 
in the cylinders in as near a natural condition as possible, not being 
worked, dried, screened or otherwise manipulated. The cylinders 
were filled to within 4 cm. of the open end as shown in figure 23. 

A 

Fig. 23. Method of placing the cells against the soil and 
holding them : S, soil ; Su, sugar solution ; M, semipermea- 
ble membrane ; C, cotton ; T, thermometer ; and R, rubber 
band passing around the cylinder and cork supporting the 
cell, thus holding the cell in place. 




After the soil had been leveled off as smooth as possible, the bases of 
the cells were slightly moistened and placed against the soil, pressed 
firmly against the surface, and rotated several times to secure a good 



338 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



contact. Two strong rubber bands (R, figure 23) were used to hold 
the cells in place and insure at all times a good contact with the soil. 
Cotton was placed over the exposed surface of the soil around the 
cells to avoid excessive evaporation (C, figure 23). A sample of the 
soil was taken at the time the cells were put in place for moisture 
determination. 

Both cylinders were filled with sandy loam, muck, and clay, re- 
spectively, in the manner described and the apparatus set up as shown 
in figure 23. The results obtained from each of the soil types are 
given in Table 2. In working with the sandy loam and clay a 5 per- 
cent sugar solution was used, but for the muck a 2 percent solution 
was found to work better. 

Table 2. — Movement of a sugar solution against the solution contained in 
sandy loam, muck, and clay soils. 



SANDY LOAM CONTAINING Ip PERCENT MOISTURE. 





Cell No. I. 


Cell No. 2. 


Time in 
hours. 




Readings on 






Readings on 






Temperature. 


graduated 


Rise. 


Temperature. 


graduated 


Rise. 






pipette. 






pipette. 






° c. 


CC. 


mm. 


° c. 


CC. 


mm. 





20 








19 








24 


19 








18 








48 


20 








18 








60 


21 


.10 


2.0 


20 


•15 


3-0 


72 


22 


.20 


4-5 


21 


•25 


6.0 


84 


22 


.25 


6.0 


21 


.35 


8.0 


96 


19 


•25 


6.0 


20 


•45 


10. 


108 


24 


•45 


10. 


23 


•65 


15.0 


MUCK SOIL CONTAINING 34-19 PERCENT MOISTURE. 





18 








18 








17 


18 


- -3 


- 7.0 


20 


- -7 


— 16.0 


41 


20 


- -4 


- 9.0 


20 


- -9 


— 20.0 


65 


22 


2.2 


50.5 


22 


— I.O 


-22.5 


89 


Out top 10 c.c. pipette 


24 


.2 


4-5 


113 








Out top 10 c.c. pipette 



CLAY SOIL CONTAINING 35.64 PERCENT MOISTURE. 






25 










25 


1 

1 





24 


22 


3.10 




72.5 


22 


370 1 


85-5 


48 


23 


8.70 




201.0 


Out top 10 c.c. pipette 




/3o 


Out top 10 c.c. 


pipette 









Discussion. 

While the above data are not very extensive and are comparatively 
simple in many respects, they present a few problems in a field in 



se ll wsTiou : A s•| ^l)^■ oi' soil solutions. 339 

which Httle, if any, work has been done. It has lonj^ l)een known 
that soil moisture responds to two forces, viz., gravitational i)ull and 
capillary attraction. In addition to these two forces a third force 
manifests itself, a force that is the result of the osmotic pressure oi 
the soil solution. In this work the effect of the gravitational pull was 
overcome by placing the membranes in a vertical position and that of 
the capillary attraction prevented by reducing the surface exposed in 
the graduated tubes. 

Work previously done with animal bladders or other semipermeable 
membranes show that when two solutions of different concentrations 
are separated by such membranes the more dilute solution always 
moves toward the solution of greater concentration. From Table 2 
it appears that the soil solution is of less concentration than the sugar 
solutions used, because there has been a movement of the soil solu- 
tion into the sugar solution. 

It will be noticed that the different kinds of soil containing different 
amounts of moisture act differently toward the sugar solutions. Re- 
ferring to Table 2, it will be noticed that the sandy loam seems to be 
able to withstand the pull of a 5 percent sugar solution better than 
the others. Whether this is due to the texture of the soil or to the 
amount and concentration of the moisture present is a problem in 
itself and can not be settled here, but it would seem that with less 
moisture present there would be greater concentration of the solutes 
in the soil and hence greater resistance to the osmotic pull of the 
sugar solution. 

The muck contained about twice as much moisture as the sandy 
loam and in working with it the rise of a 5 percent sugar solution in 
the pipettes was so rapid that no satisfactory readings for a period 
of time could be obtained. To overcome this difficulty a 2 percent 
solution was used. From Table 2 it will be noticed that there was 
some exosmosis at first, which was followed by a very rapid rise. 
The exosmosis was probably due to a poor contact of the cell with 
the soil. Later, the cell became firmed against the soil and the rapid 
rise of the sugar solution followed. If it is taken into consideration 
that with muck a 2 percent sugar solution was employed, while with 
sandy loam and clay a 5 percent solution was used, it will appear 
that the soil solution in muck is less concentrated than that of the 
sandy loam or clay, for there would 'be a greater rise of the sugar 
solution in the pipettes. This is what might be expected a priori, 
since muck is loose, porous, and spongy in its makeup and hence con- 
tains a greater amount of moisture than the other soils, and this 



340 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

greater amount of moisture will make the soil solution less concen- 
trated. 

That soil solutions are capable of developing osmotic pressure is 
not to be doubted. This work, though limited, was directed toward 
the development of a method by which the magnitude of the osmotic 
pressure of soil solutions might be measured. Just what is the 
strength of these soil solutions in terms of a sugar solution? The 
author hoped to be able to determine this by balancing a sugar solu- 
tion of known strength on one side of a semipermeable membrane 
against the soil containing a soil solution of unknown strength placed 
against the other side. When a sugar solution of known strength has 
been found that will give neither osmotic nor exosmotic movement 
on placing the membranes in contact with the soil, then the strength 
of the soil solution would be known. However, the necessity for 
frequent soaking and repairing of the membranes limited the time for 
experimental work and all that can be said is that the concentration 
of the soil solution (measured in terms of osmotic pressure) in sandy 
loam and clay appears to be below that of a 5 percent sugar solution 
and that of muck below that of a 2 percent solution. 



THE RELATION OF THE VIGOR OF THE CORN PLANT TO 

YIELD.i 

A. E. Grantham. 

Low yields of corn are generally attributed to poor seed, insufficient 
plant food, careless cultivation, or to a combination of these factors, 
but another condition in the development of a field of corn may afifect 
the yield. It is well known to close observers that considerable varia- 
tion exists in the size and vigor of the corn plants under average con- 
ditions in- the field. These differences are more noticeable when the 
plants are only a few weeks old ; as the crop develops they are more 
or less obscured. To what extent this variation in the size of the 
plant affects the yield has not been seriously considered or reported 
on. It is generally assumed, however, that the less vigorous plants 
fall below the average in yield. Not long ago the writer had the op- 
portunity of observing the behavior of a large number of plants 
through the growing season and of finally determining the actual 
yield. 

1 Contribution from the Delaware Agricultural Experiment Station, Newark, 
Del. Received for publication May 29, 1917. 



GRANTHAM : RKLATION OI' VIGOR TO \ 



During the latter half of June most corn fields present a condition 
somewhat as follows. If the corn is planted in hills, one of the two 
stalks often will he fairly vigorous and thrifty while its companion plant 
will he only half to two-thirds the size of the larger. It will surprise 
anyone to note the proportion of hills of corn in the average field that 
present this condition. The writer determined to follow up the de- 
velopment and maturity of a numher of hills of this character in a 
field where the soil conditions were uniform and marked a numher of 
hills in which a weak stalk was growing along with a vigorous one. 
The corn under observation was planted in hills 42 inches apart each 
way, two plants to the hill. Care was taken in selecting the hills 
that they were located where the stand was uniform with no hills 
missing. Of the first 50 hills selected in which there was a marked 
difference in the size and vigor of the two plants the weaker stalk 
was removed, leaving the more vigorous plant. In another 50 hills 
the strong stalk was removed and the weak one left. All of the hills 
marked were well scattered, no two being adjacent. The height of 
the remaining stalk was taken at the time the hill was thinned and the 
plant marked with a small stake bearing a number. This was done 
for each of the hundred plants under observation. The object of 
thinning was to eliminate any undue influence of one plant upon the 
other. At intervals of about eight days the staked plants were meas- 
ured to determine the rate of growth. These measurements were 
continued until September 18. The date of tasseling for each plant 
was likewise noted. The relative rate of growth for the two sets of 
plants from June 25 to September 18 is given in Table i. 



Table i. — Rate of growth in inches of strong and of weak stalks of corn 
measured on various dates from June 25 to September 18. 



Date. 


Height i 
Weak stalks. 


n inches. 

Vigorous stalks. 


Difference in height. 


Average. 


Gain. 


Average. 


Gain. 


Inches. 


Percent. 


June 25 


4-7 




9.2 




4-5 


95 








July 3 


9.4 


4-7 


18. 1 


8.9 


8.7 


92 


July 12 


20.6 


II. 2 


31-7 


13.6 


II. I 


53 


July 20 


26.9 


6.3 


42.2 


10.5 


15-3 


56 


July 27 


35-0 


8.1 


56.4 


14.2 


21.4 


61 


Aug. 5 


52.2 


17.2 


76.5 


20.1 


233 


44 


Aug. II 


66.5 


14-3 


87.9 


II. 4 


21.4 


32 


Aug. 19 


76.7 


10.2 


92.3 


4.4 


15-6 


20 


Sept. 18 


98.0 


21.3 


107.3 


15.0 


9-3 


9 



Table i shows that there was considerable difference in the aver- 



342 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

age height of the two sets of plants at the beginning and at the end 
of the growing period. The extreme difference in height, 23.4 inches, 
was on August 5, the date the more vigorous plants came into tassel. 
The average date of tasseling of the weaker plants was August 12, 
a difference of nine days in the time of tasseling. While there was 
95 percent difference in height on July 25, there was but 9 percent 
difference on September 18. This difference in the height of the 
mature plants is so small that it would hardly be noticed by the casual 
observer. 

When the corn had matured each set of plants was harvested indi- 
vidually. At the proper time for husking the two lots were husked 
and weighed separately. The weight of each individual ear was 
taken and the corn stored until thoroughly dry for shelling. The 
average weights of ears from plants tasseling at different dates are 
given in Table 2. It will be noted that the weight of the ears from 
the strong plants varies from 277 to 338 grams and from the weak 
plants from 60 to 283 grams. 



Table 2. — The average weight at husking time of ears from strong and from 
weak plants tasseling at various dates. 



Dates of tasseling. 


Strong plants. 


Weak plants. 










No. of ears. 


Average weight. 


No. of ears. 


Average weight. 






Grams. 




Grams. 


July 27 


4 


291 






August 3 


22 


338 


4 


283 


August 5 


8 


315 


2 


165 




6 


277 


8 


280 




I 


305 


3 


113 


August II 


5 


310 


11 


151 


August 13 


I 


290 


8 • 


163 


August 17 






7 


142 


August 19 






2 


60 


August 21 






3 


lOI 


Total or average 


47 


309 


48 


177 



The average weight of the ears from the strong plants is 309 grams ; 
from the weak plants, 177 grams, a difference of 74 percent. The 
table also shows that nearly half of the weak plants tasseled after 
Aug. II, while only one of the stronger plants came into tassel after 
that date. 

In Table 3 the distribution of the population of each group of ears 
with respect to weight is shown. 



GUANTIIAM : KI-.LATION OF VKiUR TO \ 



343 



Table 3. — Distribution of the population of cars from strong and from ivcak 
plants according to weight. 



Number of ears. 



Strong stalks. Weak stalks, 



Weight of ear, 
Riams. 



Number of ears. 



Strong stalks. Weak stalks 



300-350 
350-400 
400-450 
450-500 
500-550 



Table 3 shows that the strong plants produced no ears under 100 
grams in weight. On the other hand, there were 10 ears from the 
weak stalks that weighed less than 100 grams. More than half of 
the ears from the strong stalks weighed over 300 grams, while two- 
thirds of the ears from the weak stalks weighed less than 200 grams. 
The yield of dried shelled grain was 221.7 109.6 grams from the 
strong and weak plants, respectively; the weight of cob, 41.7 and 
38.8 grams. The strong plants had 19 percent of cob and the weak, 
24 percent. The yield of shelled grain from the strong plants was 
102 percent larger than from the weak. 

The results of this very brief study indicate that the weaker plants 
in a population of corn are much below the average in yield. The 
stand of plants in a field may be perfect and yet produce only an 
ordinary crop. The weak plants may be the result of environment, 
but not in this test, as each weak stalk had at the beginning a strong 
companion stalk. The weakness would appear to be inherited, the 
result of a lack of vigor on the part of the kernel. This inference 
naturally leads to the question of thinning corn. If care is used in 
thinning so that only the more vigorous plants are left in the hill a 
considerable advance in yield may be expected over haphazard selec- 
tion. It may also be advisable to plant several kernels to the hill so 
that a wider opportunity for selecting the stronger plants may be of- 
fered. Further work along this line must be undertaken to deter- 
mine fully the extent of these differences and the methods which may 
be employed to obtain the most vigorous plants. 



344 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



THE EFFECT OF DIFFERENT ROTATION SYSTEMS AND OF 
FERTILIZERS ON THE PROTEIN CONTENT OF OATS.^ 

R. W. Thatcher and A. C. Arny. 
Introduction. 

The Farm Crops Section of the Division of Agronomy and Farm 
Management of the Minnesota Agricultural Experiment Station has 
under way a series of plot studies of various crop rotations, 'both with 
and without the application of manure and of various forms of com- 
mercial fertilizers. These plots serve very well for a study of the 
efTect of the soil treatment and of clover in the rotation upon the 
chemical composition of the succeeding crops. Several different lines 
of study of this sort are in progress. The present paper deals with 
certain results which have 'been obtained with oats. 

Samples of the crop of oats from each of the several plots involved 
in the study were taken each year to the chemical laboratory and the 
percentage of dry matter and of protein which they contained deter- 
mined in the usual manner. No field data other than the plot num- 
bers accompanied the samples and it was not until the analytical 
figures were finally compiled for record that the regular effect of the 
various plot treatments upon the protein content of the oats was dis- 
covered. This efifect is so definite and so striking that it seems desir- 
able to publish the results at the present time, although the studies 
will be continued over a considerably longer period of years. 

There are a few previous, reports of results of analyses of oats to 
show the effect of fertilizers used upon the composition of the grain, 
but they are generally the results of a single season's tests, often with 
inconclusive results. Woods^ found an apparent increase in protein 
content of both grain and straw with increased application of nitro- 
gen in the fertilizer. Weibull,^ using the composition of the crop as 
an index for fertilizer requirement of the soil, concluded that since 
he found slightly increased percentages of nitrogen in the grain and 

1 Published with the approval of the Director as Paper No. 72 of the Journal 
Series of the Minnesota Agricultural Experiment Station. Received for pub- 
lication July 2, 1917. 

2 Woods, C. D. Effects of different fertilizers upon the composition of oats 
and straw. In Conn. Storrs Agr. Expt. Sta. Rpt. for 1892, p. 47-56. 

3 Weibull, M. Cooperative fertilizer experiments in Malmohus County, Swe- 
den, 1902. Ahs. in Expt. Sta. Rec, 15 : 570. 1903. 



TIIATC IUCK \- AUN\ : TllH I'RoriCIN CONTENT OF OATS. 345 



of potash in the straw from plots which were fertilized with those 
particular elements and no consistent increase in phosplioric acid in 
the <^rain from plots to which phosphate fertilizers had heen added, 
the soils were in need of nitrogen and potassium, but not of phos- 
l)horus. Pingree,"^ as a result of studies of oats grown in 1904, 
found that where nitrogen was applied alone a larger proportion of 
protein in the dry matter of the entire plant was produced than in 
any other of the soil treatments used, the proportion being distinctly 
less in the crop on the unfertihzed plot, still lower when potassium 
alone was used, and lowest of all whenever phosphoric acid was used 
(even in a complete fertilizer). Tretiakow,^ in a single experiment, 
found that an application of barnyard manure increased the protein 
content of oats from 11.38 percent to 12.81 percent. 

Lipman^ studied the efifect of the application of potassium sulfate 
and of sodium nitrate to oats grown alone and grown with peas in 
large galvanized iron cylinders in the open field and in pots in the 
greenhouse, and of other combinations of legume and non-legume 
under varying conditions, and came to the conclusion that 

1. Under favorable conditions non-legumes associated with legumes may se- 
cure large amounts of nitrogen from the latter, even though this may not be 
indicated by an increased proportion of nitrogen in the dry matter of the 
non-legume. 

2. When sodium nitrate is applied to such crop mixtures, the non-legumes 
gain an advantage in the competition for moisture, light and plant-food, and 
the growth of the legume is depressed. The latter contains not only less of 
dry matter and nitrogen, but may possess a smaller proportion of nitrogen in 
the dry matter. 

Lyon and Bizzell" have noted an increased protein content of tim- 
othy w^hen grown in association with alfalfa or with clover and of 
oats when grown with peas as compared with that of the grass or 
cereal when grown alone on adjacent plots in the same season. 

None of these studies, however, have dealt with the efifect of a 
legume in the rotation upon the protein content of other crops grown 
in intervening years upon the same plot. So far as we are aware, 

* Pingree, M. H. The influence of nitrogenous, phosphatic and potassic fer- 
tilizers upon the percentage of nitrogen and mineral constituents of the oat 
plant. In Penn. Agr. Expt. Sta. Rpt., 1906, p. 43-53. 

5 Tretiakow, S. S. F. Influence of mode of cultivation on the chemical com- 
position of cereals. Abs. in Expt. Sta. Record, 34 (1916), p. 230. 

^ Lipman, J. G. The associative growth of legumes and non-legumes. N. J. 
Agr. Expt. Sta. Bui. 253. 1912. 

Lyon, T. L., and Bizzell, J. A. A heretofore unnoted benefit from the 
growth of legumes. New York (Cornell) Agr. Expt. Sta. Bui. 294. 191 1. 



34^ JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

the data presented below are the first to be presented which deal with 
this phase of the problem. 

The crops from which the samples were taken for our analyses 
were in regular rotation series. Hence, with the exception of the 
plot sown to oats continuously, the crop which had received the 
given treatment in any given year grew upon a different plot than in 
the preceding year or, in the 4-year and 5-year rotations, in any of the 
previous years. This reduces to a minimum the possibility of the ob- 
served effects being due to soil differences or to accumulations from 
previous soil treatments. We believe, therefore, that the results here 
presented afford conclusive evidence on the points in question. 

The particular cultural and soil treatments the effects of which are 
considered in this paper are briefly described below. In all of the 
systems except where oats are grown continuously the crops are 
grown successively on as many plots as there are crops in the rota- 
tion. The same variety of oats was used for the entire series and 
the rate and date of seeding, method of harvesting, etc., were uni- 
form for the series each year. 

Rotation Plots. 

Continuous Oats. — This plot has been seeded to oats each spring 
since 1909. Manure is applied to the plot at the rate of 6 tons per 
acre every third year in the autumn preceding the preparation of the 
seed bed for the next year's crop. The last application was in the 
autumn of 1915. 

Two-year Rotation^ Oats and Wheat. — Treatment the same as for 
the plot on which oats are grown continuously, except that oats and 
wheat are alternated on two plots. 

Tzvo-year Rotation, Oats and Corn. — Same as preceding except 
alternate plantings of corn and oats. 

Three-year Rotation, No Manure. — Corn, oats, and clover are 
seeded in rotation, with no manurial treatment. 

Three-year Rotation, ''Model Rotation." — Same as preceding plot, 
except that manure is applied the autumn previous to the planting of 
the corn at the rate of 6 tons per acre. 

Four-year Rotation. — Corn, oats, wheat, and clover are seeded in 
rotation, manure being applied the autumn preceding the corn at the 
rate of 8 tons per acre. 

Five-year Rotation. — Corn, oats, wheat, and clover and timothy 
hay are grown in rotation and the plot pastured during the fifth year. 
Manure is applied the autumn preceding the corn at the rate of 10 
tons per acre. 



THATCIIKR .-v AUNV: TIIK I'kOTKIN CONTENT OF OATS. 



347 



Fkktii,i/i:r Plots. 

All of these plots are in a 3-ycar rotation of corn, oats, and clover. 
The fertilizer is applied annually. The different fertilizers arc pur- 
chased and applied separately. The phosphates and potash are ap- 
plied at seeding time. The nitrate is applied after the grain and corn 
are up. The following are the kinds and amounts of fertilizer used. 

Commercial Fertiliser Only. — Two hundred and fifty pounds of 
acid phosphate and one hundred pounds of muriate of potash per 
acre are applied, half to the oats and half to the corn at the time the 
seed bed is prepared. After the grain is up 300 pounds of nitrate of 
soda per acre are applied, half to each crop. 

Manure and Commercial Fertiliser. — Manure at the rate of 6 tons 
per acre and the same commercial fertilizer as used on the preceding 
plot are applied. 

Manure and Nitrate of Soda. — Six tons of manure per acre are 
applied to the corn each year and in addition 320 pounds of nitrate 
of soda per acre, half to the oats and half to the corn. 

Manure and Muriate of Potash. — Six tons of manure per acre are 
applied to the corn each year and in addition 200 pounds of muriate 
of potash per acre, half to the oats and half to the corn. 

Manure and Raw Rock Phosphate. — Six tons of manure and 1,000 
pounds of raw rock phosphate per acre are applied to the corn each 
year. 

Manure and Acid Phosphate. — Six tons of manure per acre are 
applied to the corn and in addition 400 pounds of acid phosphate per 
acre, half to the oats and half to the corn. 

Results of Analyses. 

The results of the analyses of the oat crops from the various 
plots are shown in Tables i and 2. 

These results clearly show a definite effect of the rotation system 
upon the chemical composition of the oat crop. The short rotations, 
with no clover or intertilled crop to provide for summer cultivation 
to the land, uniformly yield oats with a low percentage of protein. 
The 3-year rotations, with clover and either with or without manur- 
ing, with at least one corn crop to provide summer cultivation, give 
oats of medium protein content ; and the longer rotations, with clover 
or with clover and pasture, yield oats of high protein content. 

The results presented in Table 2 show a definite correlation be- 
tween the protein content of the oat grain and the fertilizer treat- 
ment. The plots receiving fertilizers which contain nitrogen invari- 



348 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Table i. — Effect of differential rotation systems upon the protein contents of 
oats, expressed as percentage of protein in the dry matter. 







Percentage of protein in dry matter. 


Rotation. 


Manure applied. 






I 16 


Aver- 






1914. 




age. 


None, continuous oats 


6 tons per acre each 3d year 


12.94 


1 1 .96 


13.02 


12.64 


2-year, oats and wheat 


do. 


12.63 


12.17 


12.73 


12.51 


2-year, oats and corn 


do. 


13-25 


*ii-95 


13-13 


12.78 


3-year, oats, clover, 'corn . . . . 


None 


14.00 


14.66 


15.46 


14.71 


3-year, oats, clover, corn .... 


6 tons per acre, preceding 












corn 


14.63 


13-45 


14.92 


14-33 


4-year, wheat, clover, corn, 












oats 


8 tons per acre, preceding 


15-25 


15-73 


14.89 


15.29 




corn 










5-year, wheat, clover, pasture, 












corn, oats 


10 tons per acre, preceding 


15-88 


14.49 


15-05 


15-14 




corn 











Table 2. — Effect of different fertilisations upon the protein content of oats 
grown in a 3-year rotation of oats, clover, and corn, expressed as 
percentages of protein in the dry matter. 



Fertilizer used. 




Percentage of protein' in 


dry matter. 




1913- 


1914. 


1915- 


1916. 


Average. 


None (check plot) 


14.56 


14.63 


13.09 


14.92 


14.30 


Commercial only 


16.00 


15-31 


13-57 


16.10 


15.24 


Manure + commercial 




14.69 


15.00 


15-57 


15.09 




15-13 


15-88 


16.14 


15.80 


15-74 


Manure + muriate of potash . . . 


13-81 


13.69 


12.06 


14.06 


13.40 


Manure + raw rock phosphate. . 


14.12 


13.69 


14.06 


14.76 


14.16 


Manure -\- acid phosphate 


14-31 


13-94 


15.10 


14.40 


14.44 



ably produce grain having a higher protein content than that from the 
plots which received any other treatment. The single sample having 
the highest percentage of protein and the highest average for the 
4-year period resulted from the use of nitrate of soda. The complete 
fertilizer contained enough readily available nitrogen to produce 
nearly the same effect upon the composition of the oats as the sodium 
nitrate alone. The potash fertilizer produced oats which were 
slightly lower in protein content than those from the check plots in 
every one of the four years. The phosphate fertilizers did not ma- 
terially change the protein content of the grain, that from the treated 
plots being sometimes slightly higher and sometimes slightly lower 
than that from the check plots, with the average protein content prac- 
tically identical in the check, the raw rock phosphate and the acid 
phosphate plots. 



ACIKONOM If Al- I\\l KS. 



349 



AGRONOMIC AFFAIRS 



THE TENTH ANNUAL MEETING. 

The tenth annual meeting- of the Society will be held in Washing- 
ton, D. C, on the date tentatively announced in the September num- 
ber (November 12 and 13). A number of titles for papers have 
already been submitted, but, there is still room on the program for a 
few more. Those who expect to attend and to present papers are 
urged to notify the Secretary at once, so that the full program may be 
printed and mailed to members in advance of the meeting. 

i 

MEMBERSHIP CHANGES. 

The membership reported in the last number was 655. Since that 
time 6 new members have been added and 8 have resigned, making 
the present membership 653. The names and addresses of the new 
members, the names of the members resigned, and such changes of 
address as have been reported to the Secretaty follow. 

New Members. 

FoERSTERLiNG, H., 380 High St., Perth Amboy, N. J. 
Gray, W. R, County Agent, Woodward, Okla. 

Jackson, L. D., Western Canada Flour Mills Co., Winnipeg, Man, 

Metzger, J. E., College Park, Md. 

Murray, James, Macdonald" College, Quebec, Canada. 

PiTTMAN, D, W., Agr. Expt. Sta., Logan, Utah. 

Members Resigned. 

Craig, C. E. McQuarrie, C. K. Voigt, Edwin. 

Garren, G. M. Schick, Geo. M., jr. Wehrle, L. P. 

KoEBER, James. Smith, Herbert G. 

Changes of Address. 

Andrew, Myron E., Station A, Ames, Iowa. 
Bell, N. Eric, Greenville, Ala. 

Bledsoe, R. Page, 210 Elizabeth St., Charleston, W. Va. 
Bryant, Ray, Frederick, Okla. 

Cutler, G. H., University of Alberta, Edmonton South, Alia., Canada. 
Hill, Pope R., Toccoa, Ga. 



3 so JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Kemp, A. R., Bridgeport, 111. 

Macfarlane, Wallace, University of Arizona, Tucson, Ariz. 

Robertson, R. B., R. R. No. 2, Box 5, Vinson, Okla. 

ScHUER, Henry W., 249 W. loth St., Columbus, Ohio. 

SouTHwicK, Everett F., 208 Lowell St., Peabody, Mass. 

Stewart, Geo., Agr. Expt. Sta., Logan, Utah. 

SuDDATH, R. O., Auburn, Lee Co., Ala. , 

WiLLARD, C. J., Dept. of Farm Crops, O. S. U., Columbus, Ohio. 

NOTES AND NEWS. 

W. A. Albrecht, formerly of the Missouri university and station, 
is now acting head of the department of agronomy in the University 
of Wyoming. 

W. W. Baer, assistant chemist of the New York State station, has 
entered the U. S. naval service. 

J. F. Barker, agronomist of the New York State station, has been 
placed in charge of extension work in agronomy at Ohio State Uni- 
versity, and has been succeeded at Geneva by R. C. Collison, formerly 
associate chemist. Mr. Collison has been granted a year's leave of 
absence for study. 

W. L. Carlyle, formerly director of the Oklahoma station, is now 
in charge of a 4,000-acre farm in the Canadian Northwest. 

D. A. Coleman has been appointed assistant agronomist of the 
New Jersey college and station. 

Frank R. Curtis has been appointed director of the Canebrake sta- 
tion at Uniontown, Ala. L. H. Moore, the former director, will re- 
main as assistant director. 

G. H. Cutler, professor of cereal husbandry in the University of 
Saskatchewan, is now professor of field husbandry in the University 
of Alberta. 

A. W. Gilbert, who spent the past year in graduate study in rural 
economics at Harvard University, has resigned his position as pro- 
fessor of plant breeding in Cornell University and is now connected 
with the Boston Chamber of Commerce. 

J. D. Harper, assistant in crops extension in Purdue University, is 
now county agent in Laporte County, Ind. 

Ralph D. Hetzel, director of extension in Oregon, has accepted the 
presidency of the New Hampshire Agricultural College and has 
entered on his duties. He has been succeeded in Oregon by O. D. 
Center, formerly director of extension in Idaho. 



ACRONOM IC AKK AIRS. 



E. R. ITod^soM, associate agronomist at the Virginia college and 
station, has become specialist in agronomy in the Virginia extension 
service, and has been succeeded in the college and station by T. K. 
Wolfe, formerly assistant agronomist. 

R. R. Hudelson, assistant professor of soils in the University of 
Missouri, is absent on a year's leave and is enrolled in the Reserve 
Officers Training Camp at Fort Sheridan, 111. 

R. A. Kinniard has been promoted from extension instructor to 
extension assistant professor of soils in the University of Missouri. 

M. W. Kirkpatrick, formerly superintendent of the Dodge City, 
Kans., substation, has succeeded C. E. Cassel, resigned, as superin- 
tendent of the Tribune substation in the same State. 

E. H. Lindley, professor of philosophy in the University of Indiana, 
has been elected president of the University of Idaho and has entered 
on his duties. 

Franklin L. McVey, president of the University of North Dakota, 
has been elected president of the University of Kentucky. 

Wallace Macfarlane, who for the past year has been assistant 
agronomist in charge of soils work at the Oklahoma station, is now 
agronomist of the University of Arizona. 

C. E. Myers has been promoted to associate professor of plant 
breeding and C. F. Noll and J. W. White to associate professors of 
experimental agronomy in the Pennsylvania college and station. 

J. S. Owens has been appointed assistant in experimental agronomy 
at the Pennsylvania station. 

T. S. Parsons, agronomist of the Wyoming university and station, 
has been granted a year's leave of absence for graduate work at the 
University of Wisconsin. 

Joe Robinson has been appointed assistant agronomist in the Wyo- 
ming university and station, succeeding P. T. Meyers, resigned to 
become county agent for Campbell County, Wyo. 

Henry W. Schuer and C. J. Willard, the latter from the University 
of Illinois, have been added to the instructional force in farm crops 
in Ohio State University. 

George Stewart, who spent the past year in graduate work in Cor- 
nell University, has been made assistant professor of agronomy and 
assistant agronomist in the Utah college and station. 

C. A. Thompson has been appointed assistant in soils and O. E. 
Barbee assistant in farm crops at the Washington station. 



352 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Donald K. Tressler, formerly assistant in agricultural chemistry 
in the Cornell station, is now with the U. S. Bureau of Soils. 

R. G. Wiggans, assistant professor of farm crops in the Ohio State 
University during the past year, is again associated with the farm 
crops department of Cornell University. 

C. G. Woodbury, horticulturist, has been elected director of the 
Indiana station. 

Third Interstate Cereal Conference. 

The third Interstate Cereal Conference was held at Kansas City, 
Mo., June 12-14, 191 7- The attendance was about 60, representing 
8 States, 8 offices of the U. S. Department of Agriculture, the Minne- 
sota and Kansas grain inspection departments, several commercial 
concerns, and the Southwestern farm and trade press. A wide range 
of topics was discussed, though much of the time was devoted to 
means of increasing and conserving cereals during the present emer- 
gency. A part of the third day of the conference was spent in the 
inspection of mills, elevators, and other points of interest in Kansas. 
City. A majority of those in attendance visited the Kansas station 
and college at Manhattan on June 15 and a considerable number con- 
tinued the trip to include the branch station at Hays on the follow- 
ing day. 

Conference of Western Agronomists. 

The second annual conference of agronomic workers in the eleven 
western States was held July 31 and August i and 2 at Moscow, 
Idaho, and Pullman, Wash. The first day's sessions and that on the 
morning of the second day were held at Moscow, and the remaining 
ones at Pullman. Seven States and several offices of the U. S. De- 
partment of Agriculture were represented, the State and departmental 
forces including men from 15 stations. The program included dis- 
cussions of fallowing, rotations, tillage methods, the use of irrigation 
water, the nitrogen supply, the distribution of superior seed stocks, 
marketing, adaptation of varieties, the use of new crops, interstate 
cooperation in experimentation, and other topics of interest to agro- 
nomic workers. The next conference will be held at Corvalhs, Ore., 
on a date to be decided by the executive committee, of which Gorge 
R. Hyslop is the local representative. 



JOURNAL 

OF THE 

American Society of Agronomy 



Vol. 9. November, 1917. No. 8. 



WHY CEREALS WINTERKILL.^ 

S. C. Salmon. 

Introduction. 

Winterkilling has been investigated as little as any of the important 
phases of cereal crop production. Practically no experiments to de- 
termine its causes have been conducted in the United States and only 
a few in foreign countries. The status of present knowledge depends 
largely on general observation and a few experiments with other 
plants, chiefly vegetables and fruits. 

Where winter cereals can be successfully grown they usually yield 
from a few bushels more than to several times as much as spring 
varieties. There are other advantages, such as early maturity, dis- 
tribution of labor, and condition of the ground for seeding. Only 
about one fourth of the world's wheat crop is sown in the fall, how- 
ever, and a much smaller proportion of oats and barley. The winter 
varieties of these grains are not sown mainly because of their inability 
to survive the severe winters of the principal grain-growing areas of 
the world. 

This paper is intended to suggest some probable causes of winter- 
killing, but especially to bring together the results of experiments and 
general observation as a basis for further investigations. 

Causes of Winterkilling. 

The probable causes of winterkilling may, for convenience, be 
grouped under four heads, though in some cases the boundaries over- 

1 Contribution from the Kansas Agricultural Experiment Station, Manhattan, 
Kans. Received for publication May 7, 1917. 

353 



354 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

lap. They are (i) heaving, (2) smothering, (3) physiological 
drought, and (4) direct effect of low temperature on the plant tissue 
and protoplasm. 

HEAVING. ^ 

Heaving is one of the most common causes. It occurs especially 
on poorly drained soils in humid areas and is believed to be more 
common on heavy than on light soils. It occurs usually in the spring 
and is due to alternate freezing and thawing, which expands the soil 
and then allows it to contract. The plants are lifted from the soil 
and the roots are broken and exposed to the air. Heaving is a 
common cause of winterkilling in the eastern half of the United 
States. 

No method is known for preventing heaving. Drainage of wet 
areas will aid but will not entirely prevent heaving, since it occurs 
on soils that are normally well drained. Well-prepared ground, 
timely seeding, and other conditions which promote a healthy growth 
and strong roots are important. 

Montgomery (43)^ found that rolling early in the spring where 
heaving had taken place prevented much of the injury that would 
otherwise occur. 

Newman and Pickett (47) report that in South Carolina plants 
from deeply sown seed are more likely to be injured than are those 
from shallow planting. The explanation is as follows : The soil 
seldom freezes deeper than 3 inches. If the grain is sown deep, the 
seminal roots are below the frost line and the coronal roots above. 
When the ground freezes, the upper part of the plant together with 
the coronal roots is lifted with the surface soil, while the seminal 
roots are held below. As a result the connection between the two 
sets of roots is broken. If the seed is covered shallow, however, 
both will be in the frozen soil and will be lifted together. 

Wollny (78) states that plants from shallow planting are less 
likely to be injured than from deep planting, but he attributes the 
difference to more vigorous plants in the former case. 

Wright (79) in experiments conducted in Indiana, found a de- 
cidedly greater survival from seeding i^^^ inches deep than from 
seeding either three fourths of an inch or from 3 to 4 inches deep. 

McClelland (36) found that winter oats sown in deep furrows 
in Georgia were protected from injury due to heaving by dirt from 
the ridges falling about the crowns of the plant. Similar results 
ivere secured by the writer at the Kansas station (57). 

2 Figures in parentheses refer to papers similarly numbered in the bibliog- 
raphy on page 377. 



salmon: wiiv ci;ki;.\ls wintilKkill. 



355 



SMOTIIKin NC. 

Sinothcrini;' is hclicxcd to l)t' a fri'(|iu'nt cause of injury wlicu the 
<;"rain is covered w itli an ice sheet or very deep snow. Whether the 
])lants (he from hick of air, as iniphed, or from some other cause 
seems never to have heen determined. 

Watson and Miller (72) note that varieties of wheat at the Penn- 
sylvania station were badly damaged by an ice sheet during early 
March, 1904. Nearly all the wheat so covered was killed. The in- 
jury was attributed to smothering. Similar damage was noted by 
Noll (49) for the winter of 1909-10. The ice sheet in this case was 
several inches thick. It occurred earlier in the season and remained 
for a longer period, i. e., during most of January and February. The 
injury was less, however, than in 1904. There was little relation 
between variety and extent of injury except that the Turkey and a 
spring wheat sown in the fall suffered more than the others. Timothy 
sown with the wheat was injured less than the wheat. 

Winter grains at Manhattan, Kans., and elsewhere in Kansas were 
badly damaged during the winter of 1909-10. The injury was at- 
tributed to an ice sheet formed by refreezing of melted snow. The 
survival w^as greatest south or east of a hedge, stone wall, or other 
obstruction to the wind. In these locations the snow which fell in 
midwinter did not change to ice so soon, the damage being greatest 
where the ice formed earliest. Early sown wheat survived better 
than late sown and apparently those varieties survived best which 
had the densest growth of leaves. 

In 1916 an ice sheet from half to or more inches thick occurred 
at Manhattan, Kans., without apparent injury to any winter grain. 
This ice sheet was caused by a storm of sleet and rain, which froze 
nearly as rapidly as it fell. The surface of the ground, the sleet, 
and the melted snow surrounding the plants was frozen in a solid 
mass. The storm occurred January 25 and the ice remained until 
February 10. Considerable injury was reported in the northeastern 
part of the state, where the ice was thicker. 

Sinz (68), among other observations relating to winterkilling of 
cereals at Goettingen, noted that an excessive snow covering may 
suffocate the plants. 

Waldron (71) says that in northern Michigan wheat is often 
severely injured by snow, and that in Hungary a continuous snow 
covering upon the winter wheat is considered injurious. If snow 
crusts are formed they are broken by stock driven over the fields. 
He attributed a part of the injury to wheat at the Dickinson (N. 
Dak.) substation in 1910 to smothering because of deep snow. 



356 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Wright (79) sprayed plots of wheat in February with a fine spray 
of water when the air temperature was — 12° to — 15° C. When 
the water was frozen the operation was repeated until the blades 
of the plants and the surface of the ground were completely covered 
with ice. This ice remained on about three days. When it had 
thawed the grain was found to be seriously damaged and in the 
spring even the roots were dead. A microscopic examination of the 
leaves showed that the cellular structure had been decidedly dis- 
arranged. In many places the epidermis was entirely separated from 
the underlying cells, but the most evident efifect was the disunion of 
the cells. In another experiment plants entirely submerged in water 
and others with their roots only submerged were frozen for 48 hours. 
Those entirely submerged were the more seriously damaged. A 
microscopic examination of the tissue showed conditions similar to 
those found in the first experiment. 

PHYSIOLOGICAL DROUGHT. 

Physiological drought has never been proved to be a cause of 
winterkilling of cereal crops, but has long been regarded as a cause of 
injury to shrubs and trees. A cold soil and especially a frozen soil, 
as noted by Schimper (65), is physiologically dry; that is, plants 
cannot obtain water from it. Jost (31), however, states that some 
plants may obtain water from a frozen soil. 

However, transpiration may take place quite rapidly at rather low 
temperatures, as shown by Wiesner and Pacher (75). Twigs of 
horse chestnut and oak, for example, lost 0.32 percent and 0.25 percent 
respectively of their weight in 24 hours at — 3.5° to — 10.5° C, and 
0.199 percent and 0.192 percent at — 5.5° C. to — 13.0° C. Beach 
and Allen (5) found a loss of from 4 to 9 percent of water in apple- 
tree twigs during a single week in January with a minimum tempera- 
ture of — 15.0° F. ( — 26° C). They found also in general that 
the hardiest varieties were most resistant to loss of water. 

Bud scales, corky integument, and especially cutinized protective 
coverings which were once thought to protect the plant by prevent- 
ing loss of heat are now regarded as means to prevent excessive trans- 
piration. Wiegand (73), for example, measured the temperature of 
buds with the bud scales removed and before they were removed, but 
failed to find any marked difference. Chandler (11) found that buds 
with the scales removed were slightly more resistant to low tempera- 
ture produced artificially in the laboratory than were the normal 
buds. Schimper (65) found that desert plants frequently have a 



salmon: why ci:Ki:ALS wi nticuk ill. 



357 



strong resemblance in their structure and haljit of growth to those 
of polar regions. This would be expected if resistance to cold de- 
pends on reducing the transj)iration to a minimum. 

Sachs (56) observed that the foliage of certain plants wilted when 
exposed to a temperature above the freezing point. He concluded, 
as did Miiller-Thurgau (45), that this wilting was due not directly 
to the cold but to the inability of the roots to secure moisture from 
the cold soil. 

Hall (25) states that winter injury to such shrul)s as rose trees is 
frequently due to drying winds when the roots are unable to secure 
water because of the cold soil. 

Experiments at the Arizona station appear to show a definite rela- 
tion between thickness of the epidermis and the amount of reserve 
or storage material in spineless cacti and their resistance to freezing. 

Perhaps the best evidence that physiological drought is a cause of 
injury to cereals is the well-known xerophytic structure of the most 
hardy types. Winter rye and Turkey and Kharkof wheat, for 
example, are characterized by a narrow leaf and a prostrate habit of 
growth. The soft winter wheats, winter barley, and the common 
varieties of winter oats, on the other hand, have broad leaves which 
usually assume a more or less upright position and hence are more 
exposed to the wind. The Winter Turf variety of oats has a narrow 
leaf and a prostrate habit of growth very similar to that of Turkey 
and Kharkof wheat. This variety is the hardiest of the winter 
oats. It is much less hardy, however, than those varieties of wheat 
and barley which are characterized by wide leaves and upright 
growth. 

Federoff (16) observed that wheat was injured much less in 
protected portions of a field than in exposed portions. The injury 
was attributed to a March wind when the ground was frozen, there 
having been no injury during the preceding severe winter. 

Kolkunov (32) found that those varieties of wheat with the most 
pronounced xerophytic characters seemed the most hardy. 

Sinz (68) concluded as a result of experiments at the University of 
Goettingen that those varieties of wheat which seemed to be able 
to prevent rapid transpiration were among those most highly resistant 
to cold. 

Schafifnit (64), on the other hand, as a result of a study of 200 
varieties of wheat during three winters, found no relation between 
morphological characters and resistance to cold. He concluded that 
structural dififerences are unimportant from this viewpoint. 



35^ JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

In a preliminary study of several varieties of winter wheat, rye, 
barley, and oats the author and assistants (6i) found no differences 
in cell structure, epidermal covering, or ability to control transpira- 
tion that could be correlated with the great dififerences in cold-resist- 
ance known to exist. Recent data, however, indicate a relation be- 
tween the ratio of root length to leaf area and ability to survive low 
temperature. The ratio for Turkey wheat, for example, was found 
to be about 25 percent greater than that for Fultz, a less hardy variety, 
and about 40 percent greater than for oats and barley. 

Schimper (65) observes that "the capacity to withstand intense 
cold is a specific property of the protoplasm of certain plants and is 
quite unassisted by protective measures that are external." " Our 
present power of observation," he says, " does not enable us to recog- 
nize in plants any special protective means against cold." 

If physiological drought is a common or the only cause of winter 
injury, then those plants which are most resistant to drought would 
be expected to be best able to withstand severe cold. This relation 
holds true for many plants, but certainly it does not for all. Brome- 
grass, for example, is markedly resistant to both cold and. drought 
but Kentucky bluegrass, timothy, meadow foxtail, reed canary grass, 
clover, and alsike clover, which are very resistant to cold are adapted 
to humid areas only. Piper (54) states that red clover is probably 
more resistant to cold than alfalfa and that alsike clover, which is 
grown only under humid conditions, is more resistant to cold than 
either red clover or alfalfa. White clover, which, according to Piper, 
occurs northward to the limits of agriculture, grows only where mois- 
ture is fairly abundant. 

Schimper (65) notes that in central Europe delicate plants like 
Bellis perennis and Stellaria media are exposed to the weather with- 
out any hairy covering or protective layers of any kind, not even a 
thick cuticle. They are frozen hard and brittle as glass, but when 
spring comes they continue to grow undisturbed. 

In conclusion, the experimental and observational evidence does 
not permit one to deny or affirm that physiological drought is a 
cause of injury in all cases. One familiar with winter grains in the 
Great Plains of the United States can hardly escape the conviction 
that exposure to cold, dry winds when the ground is bare and frozen 
is very injurious. On the other hand, in certain northern areas winter 
wheat is usually killed close to the ground, so that not a spear of 
green can be seen. In the spring the plants start growth apparently 
as vigorously as though the green leaves had been retained. 



SALMON : Win' ('I':ui:als vvintickkill. 



359 



Also, there seems to be little evidence to show that the well-known 
difference in hardiness of winter oats and barley on the one hand and 
wheat and rye on the other is dne to the j;reater ability of the latter 
to control transt)iration or to absorb water from a cold soil. P)arley 
and oats are killed mnch more easily than wheat and rye when covered 
with snow or protected from transpiration in other ways. 

Probably, easily injured plants are killed l)efore physiological 
drought can have any marked effect. Plants which are able to 
survive temperatures that freeze the ground as deep as the roots 
penetrate, however, must frequently be exposed to the effect of phys- 
iological drought. Any habit of growth or structure which enables 
a variety to reduce the transpiration in proportion to the water ob- 
tained from the soil or any character or quality which permits it to 
survive with less water would presumably prolong its life in com- 
parison with those varieties which lack this ability. 

DIRFXT EFFECT OF LOW TEMPERATURE. 

There can be no doubt that plants are often killed by the direct 
effect of cold on the tissue without heaving, smothering, or physio- 
logical drought taking place. But the final effect of a given tempera- 
ture is modified and influenced by so many factors both external and 
internal that the nature and the exact cause of such injury is difficult 
to determine. In general, it may be attributed to one or more of 
four groups of factors, viz., (i) mechanical injury, (2) desiccation 
of the protoplasm, (3) chemical effects, and (4) suspension of meta- 
bolism. It is evident that two or more of these results may occur 
in the same tissue at the same time. 

Mechanipal Injury. — Early observers believed that plants were able 
to develop heat and so prevent the formation of ice. Early Greek 
philosophers who were unaware of the cellular structure of plants 
thought the injury to be due to the rending and mashing of the plant 
organs by the formation of ice. Du Hamel and Buffon presented 
the theory in 1737 that death was due to rupturing of the cell walls. 

Geoppert (20) found that ice formed both within the cells and 
in the intercellular spaces. Mitller-Thurgau (45) decided as a 
result of careful study that ice usually formed in the intercellular 
spaces, and within the cells only in case of rapid freezing or in 
exceptionally large cells. Miiller-Thurgau also proved that in some 
cases the formation of ice was the cause of death, since certain plants 
when supercooled were not injured, but w^ere killed if ice formed at 
the higher temperature. 



360 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Wiegand (74) has shown that the formation of ice may separate 
dififerent tissues. In leaves rich in water the ice fuses into a sheet, 
completely separating the upper layers from the lower. In twigs 
the outer layers may be entirely separated from the inner. How- 
ever, he found in the case of hardy plants that the separation of the 
cells by the ice masses ordinarily causes no injury. 

Gassner and Grimme (18) attribute injury to green plants to volu- 
metric alterations rather than to any peculiar effect on the plant cells. 

Others attribute injury to the evaporation of the water frozen in 
the intercellular spaces before it can be absorbed by the cells. The 
injury in such cases is regarded as a result of the thawing rather than 
of freezing and is severe in proportion to the rate of thawing. If 
the tissue is thawed slowly the water returns to the cells and they 
regain their turgidity without injury. If they are thawed rapidly, 
much of the water is evaporated before it can be absorbed. The pro- 
tection afforded by spraying tomatoes, chrysanthemums, and other 
tender plants when frozen as often practised by gardeners is probably 
explained in this way. 

There thus appears to be good evidence that death in many plants 
is a direct result of mechanical injury caused by the ice. In others, 
especially the more resistant species, ice may form without apparent 
injury. Schimper (65) notes, for example, that alpine plants "while 
in blossom pass the nights in a completely frozen state and during 
the daytime are exposed to the most intense insolation " and states 
that perennials of the temperate and cold zone " may be frozen into 
lumps of ice without dying." 

Desiccation. — According to the most commonly accepted theory, 
winterkilling is due to desiccation of the protoplasm when the water 
is withdrawn and frozen in the intercellular spaces. This theory ex- 
plains the injury much the same as physiological drought, except that 
in the latter case water permanently leaves the plant tissue, while in 
the former the protoplasm loses its water, although it is retained in 
the plant tissue as ice. Physiological drought can cause death only 
when the soil is very cold or frozen, and theoretically may do so with- 
out actual freezing of the plant tissue. Death from desiccation of 
the protoplasm may occur regardless of the temperature of the soil, 
but only when the tissue is frozen. 

Wiegand (74) remarks that "it seems likely that death from 
freezing is usually if not always due to drying out of the protoplasm 
beyond its critical water content. Miiller-Thurgau (45) determined 
the amount of water withdrawn from the cell and frozen in the inter- 



salmon: \vll^■ ci:kkai.s vvi nti: kkill. 



361 



cellular spaces and found that 63.7 ])(m-c(mU was so withdrawn at — 13° 
and 79.2 percent at — 15.2° C. liotli lie and IMolisch hold tliat death 
is due to the withdrawal of water from ihe cells. 

Adams (2) suhjected dry and moist seeds of peas, barley, flax, 
turnips, red clover, meadow^ fescue, and timothy to the temperature 
of liquid air. Germination of the moist seed was greatly reduced, the 
injury being attributed to withdrawal of water from the cells and 
freezing in the intercellular spaces. 

Schaffnit (64) studied the effect of low temperature on the cell 
sap, chemical constituents, enzymes, physical changes, and death points 
of green plants, spores, and pollen grains. He divided them into 
three groups with respect to their ability to survive low temperature 
and to withstand desiccation. The first group are those for which 
water is absolutely essential, the second group can withstand a cer- 
tain degree of desiccation, and the third can stand complete drying. 
For the first group the abstraction of water is regarded as the primary 
cause of death and chemical and physical changes secondary factors. 
It is claimed that temperatures near the freezing point produce in 
some plants chemical products which represent a transition from the 
less stable to the more stable forms. The conclusion is reached that 
for a given temperature " death results from vital reactions called 
forth by the external conditions." 

Chemical Effects of Cold. — The injury to plants from low tempera- 
tures have long been regarded as possibly due to chemical changes 
taking place in the protoplasm. Abbe (i) in his review in 1895 
notes that the chemical changes brought about by frost are thought 
to be of principal importance. He states that Kunnisch likens the 
chemical change in frozen sap to freezing out of cryolites from ordi- 
nary solutions at specific temperatures. 

Gorke (21) found that certain proteids are precipitated when plants 
are frozen and apparently those plants are most easily killed whose 
proteids are precipitated at the highest temperatures. He found, for 
example, that in the easily injured begonia a denatured precipitate of 
proteid is obtained at — 3° C., while in winter rye the proteid is 
precipitated at not less than — 15° C. and in pine needles at not less 
that — 40° C. Gorke also found that the acidity of sap increases on 
cooling, which he believed aided in the precipitation and denaturing 
of the proteids of the cell. He presents the theory that death may 
be due to this precipitation and to the denaturing effect which prevents 
the reabsorption of the proteids on thawing. The precipitation of 
the proteids was believed to be due to the greater concentration of 



362 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

the cell sap as the water is extracted and frozen in the intercellular 
spaces, it being well known that certain proteids are precipitated in a 
concentrated salt solution. 

Schaffnit (62) found that proteids of rye grown in the open at low 
temperatures are not readily precipitated by freezing, while the pro- 
teids of rye grown in the greenhouse at a high temperature are readily 
precipitated. He concluded that the precipitation of the proteids is 
the only way in which the loss of water from freezing kills plant 
tissue. 

Lidforss (34) found that tender seedlings placed in a sugar solu- 
tion for a time were able to survive several degrees lower tempera- 
ture than seedlings not so treated. 

Chandler (11) found a decided protective effect from treating 
seedlings of tomatoes, cabbage, cowpeas, kale, lettuce, and the buds 
and blossoms of apples, peaches, and cherries with sugar and glycer- 
ine solutions. He argues against the theory of Gorke, however, since 
certain salts which most readily prcipitate proteids did riot reduce 
the hardiness of plant tissue when absorbed by it, and in the case of 
zinc sulfate seemed to increase the ability to survive low tempera- 
tures. Also, he found, no precipitation of proteids in the sap of 
tender twigs of apple, plum, or pear even in early autumn. Some 
slight evidence was secured that proteids in sap from greenhouse 
plants of cabbage, tomato, kale, lettuce, and peas were precipitated 
by low temperature. Chandler concluded that precipitation of the 
proteids does not explain death. 

Lidforss (34) gives an interesting explanation of the non-precipita- 
tion of proteids in resistant plants as a result of investigations con- 
cerning the winter green flora of southern Sweden. Of many plants 
belonging to a variety of ecological types, none appeared to possess 
any obvious protection against the effects of low temperature. This 
especially was found true of many delicate herbaceous annuals such 
as Holosteum, Cerastium, Lamium, Veronica', Senecio, Viola, 
Fumaria, etc. In all, however, the starch which was contained in 
their tissue changed to sugar on the approach of winter, and again 
changed to starch on the approach of spring. The sugar by increas- 
ing the concentration of the sap reduced the transpiration and the 
freezing point of the sap, and prevented the precipitation of the 
proteids. 

Maximow (40) conducted extensive experiments to determine (i) 
if the death point of plants depends on the plant structure alone 
or if it varies with the physico-chemical condition of the plant; (2) if 



salmon: CI'^KIOALS VV I N'll'.KK ILL, 



there is a cHniinishiiii^ of the death point as a result of introchiciiig 
various substances into the plant cell, and (3) the relative protective 
value of different substances. lie concluded that (i ) the introduc- 
tion of neutral substances such as alcohols, suj^^ars. and sails may 
considerably increase the cold resistance of the cells; (2) the pro- 
tective action of the solution can not be explained alone by the de- 
]:)ression of the freezing point, since the resistance to cold always in- 
creased more rapidly than this depression; (3) the degree of protec- 
tion is closely related to the eutectic point of the introduced solution, 
substances with a high eutectic point showing no protective effect ; 
and (4) isotonic solutions of different substances with low eutectic 
points possessed nearly the same protective action. Additional ex- 
periments show^ed no relation between the rate of penetration of the 
protective substance and the degree of protection. From this it was 
concluded that protection depends on the solution reaching the outer 
layer only of the protoplasm. 

In other experiments Maximow (39) concluded that killing by 
cold is probably due not simply to low temperature as such (implying 
a specific minimum temperature) but to physico-chemical changes 
set up in the plasma colloids during the formation of ice. 

Metabolism at Low Temperature. — Molisch (42) has show^n that 
plants continuously exposed to a temperature too low for normal 
metabolism but above freezing will eventually die. Hilliard et al. 
(28) state that bacteria may be killed by continued low temperature 
because of its interference with metabolism. Yeast, bacteria, and 
certain molds, according to Blackman (8), can withstand prolonged 
exposure to the temperature of liquid hydrogen. 

The fact that death may result from continued cold above those 
temperatures which usually cause death may have a practical applica- 
tion, since weak plants probably will succumb quicker than others. 

Green and Ballou (22), for example, in a study of injured and 
uninjured orchards on Catawba Island and the peninsula of eastern 
Ottawa county in Ohio following the severe winter of 1903-04, found 
that where the vitality of the trees had been lowered by anv cause, 
such as low fertility or poor physical condition of the soil, San Jose 
scale, leaf curl, peach tree borers, etc., the injury from cold was 
increased. 

Chandler (11) found that the sap of peach trees in poor nutritive 
condition has lower concentration and freezes at a higher tempera- 
ture than sap from vigorous, healthy trees. 



364 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

Conditions Which Modify the Degree of Injury. 
A serious difficulty in discovering the cause of winterkilling and 
in securing varieties and devising methods of culture to prevent it is 
the presence of numerous secondary effects which cloak or hide the 
primary causes. Some of these factors are external and some reside 
within the plant itself. Among the most important external factors 
are (i) duration and intensity of cold, (2) rate of freezing and 
.thawing, (3) protection by (a) snow, (b) mulches, (c) vegetative 
cover, and (d) uneven surface of the ground, (4) moisture content 
and kind of soil, and (5) habit of growth of the plants. Of the 
internal factors the following are worthy of mention: (i) Moisture 
•content of tissue, (2) dormancy, (3) concentration of sap, (4) size 
of cells, (5) means for controlHng transpiration, and (6) age of 
plants or maturity of the tissue. 

duration and intensity of cold. 

A common conception is to think of death resulting when a certain 
specific minimum temperature is reached, this minimum depending 
mainly on the kind of plant. This view is held by Mez (41) who 
says that " each plant has its specific minimum point at which death 
occurs due to the direct effect of cold." Chandler (11) holds that 
the work of Miiller-Thurgau, Voitlander, Maximow, and his own 
work at the Missouri station entirely refutes this conclusion. 

Pfeffer (55) points out that the intensity of any agency required 
to produce a fatal effect depends not only on its duration but also 
on many variable circumstances. Hence, the minimum points can 
only be approximately and conditionally determined. 

Schimper (65) says that ''every plant can live only at a tempera- 
ture lying between two extremes termed respectively the upper and 
lower zero points. Overstepping of these limits sooner or later, 
but at the latest within two or three days results in death." He ob- 
serves that the zero points vary for different species and also for 
different functions of the same species, and that as far as known 
at no place on the earth's surface is the temperature so low that no 
plant can withstand it. 

Jost (31) quotes experiments of Brown and Escombe and of 
Thiselton-Dyer showing that certain seeds and spores are not killed 
when subjected to — 200° C. for five days or to — 250° C. for a 
shorter period. 

Wiegand (74), from a review of the literature and his own work, 
concluded that there seems to be little if any evidence that death is 



salmon: WI1\ CI.KI.AI.S VVI NTI'-RKILL. 



due to shock or over-stinnilalioii or any olhcr action of cold wliich 
might produce the so-called cold rigor. 

McFayden (37) exposed photogenic l)acU'ria to a tcnipcialurc of 
— 190° C. for 20 hours without inij)airnient of the vit.ality or func- 
tional activities of the organisms. Phosphorescent organisms become 
non4uminous when subjected to the tem])erature of liquid air l)ut 
regained luminosity with unimpaired vigor when thawed. The same 
results were secured when they were subjected to the temperature of 
liquid air for seven days, and also when subjected to the temperature 
of liquid hydrogen. 

Wheat kernels previously soaked for 48 hours in water warm 
enough to promote germination were exposed by Wright (79) to a 
freezing temperature produced by salt and ice. Lots were removed 
at intervals of two days, dried, and then sown in well-prepared garden 
soil. There was an almost constant decrease in germination with 
length of exposure and increase in the length of time required for the 
plants to appear above ground. On the average, each two days 
freezing decreased the germination 3.43 percent. 

Schafifnit (63) found that the injury to wheat plants increased with 
duration and lowering of the temperature. 

All investigators appear to agree that for plants that may be 
killed by cold the injury increases with the degree of cold and its 
duration. 

RATE OF FREEZING. 

Perhaps no phase of injury from cold has been the cause of more 
contradictory evidence than the relative effects of slow and rapid 
thawing and freezing. 

Winkler (77) reported that leaves of evergreens and twigs of other 
trees can endure from four to six times as much cold if the change 
in temperature is gradual than if they are suddenly subjected to cold. 
Winter twigs cooled rapidly to — 22° C. were killed, but if kept for 
three days at — 16°, two days at — 18° C, three days at — 20° C, 
two days at — 22° C, three days at — 25° C, and 12 hours at — 30° 
to — 32° C. they were not all killed. 

Chandler (11) found in an extensive series of studies at the Mis- 
souri station that the rate of temperature fall is very important, espe- 
cially for winter buds. He found for example that apple buds may 
be frozen rapidly enough with salt and ice so that practically all will 
be killed at 0° F. ( — 18° C.) or slightly above, while they will with- 
stand a temperature of — 20 to — 30° F. ( — 29 to — 34-5° C.) if 
the fall in temperature is gradual. It was found that a rapid fall in 



366 JOURNAL OF THE AM.ERICAN SOCIETY OF AGRONOMY. 

temperature was more injurious if it occurred during the first part 
of the freezing period than if it occurred later. Chandler found that 
the hardiness of peach buds in a dormant condition is greatly increased 
by continuous cold preceding the date when the temperature goes low 
enough to kill. In his opinion, this capacity to withstand a low temper- 
ature appears to be due to the slow fall in temperature rather than 
to hardiness developed as the result of exposure, 

Hilliard et al. (28) found that an abrupt fall in temperature had 
a greater germicidal effect on bacteria than a gradual fall in 
temperature. 

Pfeffer (55) concluded from a review of the literature that "re- 
sistant plants withstand rapid and slow cooling equally well and 
it is doubtful whether a rapid fall of temperature is more injurious 
to plants killed by freezing than is gradual cooling. 

RATE OF THAWING. 

Sachs (56) held that the amount of injury was determined by the 
rate of thawing. Miiller-Thurgau (45) showed that the methods 
used by Sachs were inaccurate and in a large number of experiments 
was unable to detect any difference in injury due to rate of thawing 
except in the fruits of apple and pear. 

Molisch (42) secured similar results in a large number of ex- 
periments. 

Chandler (11) found that a lower temperature was required to 
kill leaves of lettuce if they were thawed slowly than if thawed 
rapidly, but in the case of all other tests including unripe pears and 
apples there was no indication that the rate of thawing had anything 
to do with the amount of injury. 

Haberlandt (24) states that living cells of perennial twigs and 
other hibernating organs must be protected from violent fluctuations 
in temperature and especially from the effect of sudden thawing. 

Wright (79) exposed barely germinated wheat kernels and wheat 
plants about three weeks old to a temperature of — 17° C. over 
night. A part were then placed in a room at an average tempera- 
ture of 21° C. and a part in an outdoors cold chamber to thaw slowly 
with the first moderation of weather. The germinated kernels kept 
at room temperature showed no signs of life but perished in a few 
days. Those which were placed in the cold chamber and allowed to 
thaw slowly showed themselves to be in a growing condition after 
thawing." 

In another experiment partly germinated wheat was placed in a 



salmon: wirv ckrkals wintkuk ill, 



frcozinj;- box for periods varying from J to 36 days. 'I'lic average 
decrease in germination for each 2-(lay j)eri()(l was 3.43 ])erec nt when 
the seeds were thawed somewhat slowly, and 7.59 ])ercent when 
thawed rapidly. 

Wright also reports some experiments of Tantphons in wdiich 
partially germinated wheat that was frozen and then thawed slowly 
germinated 86 percent, while that similarly treated but thawed 
rapidly germinated only 18 percent. 

Garcia and Rigney (17) found in a study of fruit trees in New 
Mexico -that if a temperature of 25° or below occurred one to two 
hours before sunrise the damage was great but if the minimum 
temperature occurred near midnight and then gradually rose to the 
freezing point so that the frozen parts had time to thaw before sun- 
rise the injury was insignificant. 

Wiegand (73) studied the effect of alternate thawing and freez- 
ing on hardy buds and twigs of several species of shrubs and trees. 
He says : " thawing seems not to harm these tissues in the least, no 
matter how frequently or how abruptly it is done. I have often 
tried the experiment of transporting twigs abruptly from — 18° to 
the w^arm laboratory (21° C.) and back several times, thus alternately 
thawing and freezing them. No matter how many times this was 
repeated no injury could be detected in the buds even when sub- 
sequently placed in the greenhouse to grow. 

The conflicting results indicate that some important factors are 
involved which the experimental methods have not controlled or 
eliminated. As a working basis it may be suggested that slow freez- 
ing may decrease the injury by (i) preventing the formation of the 
ice within the cells, (2) by giving the tissue an opportunity to dry 
out, and (3) by permitting the protoplasm to adjust itself to the 
new conditions. Slow thawing can reduce the injury only on the 
assumption that death occurs after the tissue begins to thaw. The 
only theory so far advanced that accounts for winterkilling in this 
w^ay is by permanent loss of water from the protoplasm. In such 
cases slow thawing would permit water to be reabsorbed which if 
thawed rapidly would be evaporated into the air. 

Under natural conditions, the drying out of the plant tissue such 
as occurs in the fall and early winter would appear to be especially 
important since, as shown later, the moisture content of the tissue is 
one of the most important factors which determine the injury. 



368 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



PROTECTION. 

Protection of the plants from low temperature and wind during 
the winter is universally recognized as a most potent factor in pre- 
venting winterkiUing. Lyon et al. (35) quote figures collected by 
Boussingault which show a difference in temperature of from 1.5° C. 
to 8.5° C. due to the snow covering. The writer (57) found a 
maximum difference of 26° F. (14.5° C.) at a depth of i inch in 
the bottoms of furrows and on the surface, most of the difference 
being due to the snow. 

Bouyoucos (7) compared a bare soil and one covered with vegeta- 
tion and snow. Observation for four years showed that in excep- 
tionally cold weather the temperature of the protected soil may be 
25° F. (14° C.) higher than the bare soil at a depth of 3 inches. He 
concluded that a cover of vegetation is one of the most efficient and 
expedient means of protecting the soil from low temperature during 
the winter. Fie attributed the effect to (i) arresting cold air cur- 
rents which come in contact with the bare soil and (2) to air spaces 
formed by the vegetation, which are poor conductors of heat. 

Delwiche and Moore (13) found that a cover crop in an orchard 
decreased the depth of freezing at least one half. 

Carleton (10) claims that tillering bears an important relation to 
cold resistance because the larger number of culms per unit area 
permit only a minimum exposure of each to the weather. 

On the other hand, in certain rate-of-seeding tests a larger per- 
centage of plants survived with thin than with thick seeding. The 
writer (60), for example, found in tests conducted for the U. S. 
Dept. of Agriculture at Newell, S. Dak., that winter wheat sown at 
rates of 2 and 4 pecks per acre survived with a perfect stand, while 
4.3 percent of the plants were killed when the rate of seeding was 
6 pecks and 19.4 percent were killed when the rate of seeding was 8 
pecks per acre. Again, in experiments at the Kansas station (61) in 
which wheat was sown in furrows at the rates of 4, 6, 8 and 10 pecks 
per acre, the winterkilling increased with the rate of seeding. The 
percentage survival was 75.4 for seeding 10 pecks per acre, 91.0 for 
8 pecks, 91.3 for 6 pecks, and 92.8 for 4 pecks per acre. Hume et al. 
(29) report a fair survival of winter wheat at Eureka, S. Dak., when 
2 and 3 pecks per acre were sown, while that sown 4 and 5 pecks per 
acre was entirely killed. 

Oskamp (50) recorded the temperature at a depth of 9 inches of 
three plots differently treated. The first was cultivated during the 
summer and sown to rye in the fall for a cover crop. The second 



salmon: win ci:ki:ai.s vvi nti;ukii.l, 



And tliird were left in sod. the j^rass heiiij^ ent and allowed to lie. 
Idle third ])lot was nnilehed with straw at the rale of alxinl 15 tons 
per aere. lie eonelnded "that a system of elean enltivation with a 
winter eover erop is eharaeterized hy extreme diurnal lluel nations in 
temperature ; that a straw nuileh etjualized these fluetuations to a 
marked extent, as does also a grass erop, thoui^h in a less de,i^ree." 

Wright (79) conducted experiments at Purdue University in which 
the effect of a 2-inch covering of straw, a layer of straw beneath the 
surface, and of manure applied as a top dressing or incorporated with 
the soil before seeding was studied. The straw, whether apolied as 
a surface mulch or beneath the surface, proved decidedly injurious, 
but the manure applied either as a mulch or mixed with the soil was 
beneficial. The wdieat w^as sown late and winterkilling was very 
severe. 

Green and Ballou (22) found a marked contrast in the extent of 
injury to orchard trees on bare and covered soils. The bare soils 
froze deeper and the injury was much greater than on soil covered 
with a mulch or other material. In an experiment a plot on which 
the sod was removed froze to a depth of 18 inches while a plot with 
a thin sod covering of grass and weeds froze to a depth of about 
8 inches only. 

Hume et al. (29) found that 3 tons of straw per acre spread 
on winter wheat as a mulch late in November prevented wdnter 
injury, while that not mulched was entirely killed. 

Babcock (4) sowed rows of wheat at the Williston, N. Dak., sta- 
tion in standing corn, on bare ground, and in grain stubble, a part 
of the rows being covered lightly with straw. Whenever the grain 
was protected in any way by a covering of snow or straw the winter- 
killing was very slight or none at all. Wherever the rows remained 
without covering most or all of the plants were killed. 

Bouyoucos (7) compared the temperature of a bare soil and one 
covered with straw. At a depth of 7 inches the former froze De- 
cember II and the latter February 14. At times the temperature of 
the covered plot was as much as 10° 'F. warmer than the bare plot. 

Hickman (27) compared the effect of a light straw mulch not more 
than half covering the ground, a medium mulch which covered the 
ground, and a heavy mulch from 2 to 2>4 inches thick. The effect 
on winterkilling was not recorded, but the heavy and medium mulches 
reduced the yields considerably. It is stated that the light mulch 
may have been of some value. 

Plumb mulched two wheat plots in January, spreading the straw 



37C> JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

about 3 inches deep in the loose condition. Two unmulched plots 
similarly treated in other respects produced about a bushel more 
grain than the mulched plots. The effect on winterkilling is not 
recorded. 

Mulching with 2 inches of straw failed to prevent winterkilling 
of wheat at the Indian Head, Manitoba, Experimental Farm in 1890. 
The only difference observed was that the mulched portions of the 
field remained green a few days longer than the others. 

Clark (12) reported an average survival of 33 percent for several 
varieties of winter wheat at the Dickinson, N. Dak., substation, when 
sown on corn ground with the stalks left standing as compared with 
a survival of 19 percent when sown on fallow. The average yields 
were 13.5 bushels for the corn ground and 4.7 bushels for the fallow. 
The difference was attributed to the standing stalks catching the 
snow and protecting the plants during the winter. 

Hume et al. (29) secured a yield of 28.7 bushels of winter wheat 
sown on corn ground as compared with 19.2 bushels when sown after 
oats. Winter rye after corn produced 41.9 bushels per acre as com- 
pared with 26 bushels after rye. 

In another experiment wheat sown in " narrow troughs with a 
double disk drill " came through a severe winter in good condition 
while the wheat on another field harrowed quite level after seeding 
was entirely killed. Additional data regarding the protection afforded 
by seeding in furrows is fully discussed by the writer in another 
paper (56). 

Neveroff (48) reported the effect of rolling a field covered with 
snow and then plowing so as to leave the snow in ridges to catch 
more snow. He found that the compacted snow melted slower in 
the spring and so prevented the crop from starting so early and 
being injured by following cold weather. The highest gain for this 
treatment in eleven years with winter rye was 1,060 pounds per acre 
and the least gain 84 pounds per acre. The treatment was also 
found to increase the yield of oats grown after the rye. In another 
experiment plowing the snow without rolling increased the yield of 
clover from 320 to 520 pounds per acre and alfalfa from 315 to 
915 pounds per acre. 

The author has pointed out in another paper (59) the surprising 
fact that in North America winter wheat does not appear to be able 
to survive lower winter temperatures in areas of heavy snowfall 
than in those where the snowfall is normally light, the difference if 
any being in favor of the latter locations. Smith (69), in a study 



SAi.MON : win' ci'.Ki'.Ai.s wi nti-:rkiij.. 



371 



of the relation of nictorolo}^ical factors to the production of winter 
wlieat in ( )hi(), found on the average no heneiit from a snow e(jvering 
or daniam' from laek of it. A snowfall in January appeared to be 
favorable but a snowfall in March was decidedly detrimental. Hicse 
conclusions are so opposed to common belief that a brief di.scussion 
seems called for. As pointed out in the article referred to above the 
failure of snow to protect is probably in part due to saturatin!L( the 
soil with water in the spring- and so causing more damage from heav- 
ing. It may also be explained in part by the fact that a wet soil 
when frozen is often colder than a dry soil, as shown in another 
part of this paper. A snow early in the winter would also ])e more 
effective than later since in the former case it would aid in holding 
the heat still present in the soil while later in the winter most of the 
heat w^ould have been lost. 

KIND OF SOIL AND MOISTURE CONTENT. 

The kind of soil and its moisture content undoubtedly have an im- 
portant relation to winter injury. Hunt (30), for example, says that 
the loamy soils of the corn belt which are usually friable and well 
supplied with organic matter are not so well adapted to wheat as the 
clay upland soils, because on the former wheat is likely to winterkill 
in unfavorable seasons. Montgomery (44) says that all heavy soils 
in humid regions heave, due to alternate thawing and freezing. 
Carleton (9) states that the black waxy soil of north central Texas 
is so stiff and heavy that it cracks and heaves badly in the spring, 
thereby exposing the wheat roots to the weather. Bouyoucos (7) 
found that a peat soil thawed 8 to 10 days later in the spring than 
others, due to the large quantity of water it contained. 

Petit, according to Patten (52), found that the passage of frost 
into the ground is fastest for quartz sand, slower for clay, and slowest 
for humus (peat). "For continued frost the soil temperature sinks 
after freezing faster and deeper, the lower the moisture content of 
the soil, and conversely for thawing of the soil." 

The author (57), in experiments conducted at the Kansas station, 
found that the temperature of wet clay and loam fluctuated less and 
the daily minimum temperatures were higher than for dry soil of the 
same kind until both were frozen. After they were frozen, how- 
ever, the fluctuation in temperature was usually greater and the 
minimum temperatures lower in the wet soil. The temperature of 
sand fluctuated more than dry clay or clay loam and as a result was 
the coldest as measured by the daily minimum temperature. A dry 



3/2 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

sand was found to be colder than wet sand whether frozen or un- 
frozen. The winterkilHng of grains sown on these soils followed 
very closely the average daily minimum temperatures. 

HABIT OF GROWTH OF PLANTS. 

The relation of habit of growth to winter hardiness has been dis- 
cussed, as far as its relation to physiological drought is concerned. 
But it is quite likely that plants with prostrate leaves are subjected to 
less extreme temperatures than those which are upright and fully 
exposed to the air. This especially is true when the ground is covered 
with a light snow or where there is a heavy vegetative growth. The 
greater resistance to cold of Turkey and similar types of wheat as 
compared with soft varieties may be due partly to this fact. 

MOISTURE CONTENT OF TISSUE. 

The moisture content of the plant tissue is among the most im- 
portant of the internal factors which influence winter hardiness. 
Sinz (68) concluded from experiments conducted at Goettingen that 
different varieties of wheat show a graduation in the amount of dry 
matter which is in direct relation to their resistance to low 
temperature. 

Shutt (67) determined the moisture content of twigs of fruit trees 
at Ottawa, Canada. He concluded that the data gives " direct and 
definite proof that there is a distinct relationship between the moisture 
content of the twig and its power to resist the action of frost, and 
that those trees whose new growth contain the largest percentage of 
water as winter approaches are in all probability the most tender." 

Schaffnit (63) notes that resistance to cold of some varieties of 
wheat seems to show a relation to the water content, an increase 
rendering the plant more sensitive to outside influences. Detmer 
(15) noted that wheat loses its vitality when frozen apparently in 
proportion to its moisture content. 

La Tourette (33) in experiments conducted under the direction 
of the writer found a direct relation between the moisture content 
of ungerminated wheat and resistance to low temperature. In nearly 
all cases the grain containing the most water suffered the greatest 
injury. 

Wiggans (76) soaked seeds of various kinds in water for three 
hours. Each kind was then divided into three lots, one of which 
was germinated immediately, another frozen for twenty-four hours 
at a few degrees below 0° C. and germinated, and the third lot 



SALMON : wjiv ci:ki:ai.s winti.kkill, 



373 



frozen wilh the second, treated with ether and ^^crniinated. The 
freezing rechiced the i^erniination markedly in all cases hnt the injury 
was less when followed hy etherization. The average germination 
of the seeds which were soaked hut not frozen was 60 percent, of 
those that were frozen hut not etherized 24.6 percent, and of those 
wdiich were frozen and then treated with ether 28.6 percent. 

Chandler (11) found hut little difference in moisture content of 
unfrozen cortex in seasons when it is very tender and in seasons 
when it is very hardy. De Candolle ( 14) and Picet (53) suhjected dry 
seeds to a temperature of — 80° without injury, while seeds swollen 
in water were killed at a much higher temperature. Adams (2) 
found that seeds containing less than 12 percent of water were unin- 
jured w^hen exposed to the temperature of liquid air, but those seeds 
which were moist wdien frozen were practically all killed. 

Becquerel (6) exposed seeds of castor beans, pine, squash, buck- 
wheat, corn, \vheat, oats, beans, lupines, peas, vetches, alfalfa, and 
radish to the temperature of liquid air ( — 185 to — 192° C.) for 130 
hours. It was found that the ability to withstand the low tempera- 
ture depended on the amount of water and gas present. In the 
case of moist seeds cold disorganized the protoplasm and nucleus, 
making germination impossible. But if the protoplasm has reached 
its maximum concentration by drying and at the same time its 
minimum activity the low temperature was not injurious. 

DORMANCY. 

A condition closely related to moisture content of the tissue and 
probably not less important in its relation to winterkilling is 
dormancy. 

' Chandler (11) found that the most important feature affecting 
hardiness of fruit trees is maturity of the tissue. He states that in 
the peach-growing districts of southern Missouri and in Arkansas 
the most important factor that detetmines the loss from low tempera- 
ture is warm periods during the winter which start the buds into 
growth. It is said that killing of peach buds in Connecticut, New 
Hampshire, and the peach regions of Canada generally occurs when a 
cold period follows a thaw. 

Selby (66) attributed the injury to orchard trees and shrubbery 
in Ohio in 1906-7 largely to high rainfall and temperature during 
the preceding fall, which induced late growth and high moisture 
content of the tissue. Allen (3) found that those varieties of apple 
trees which mature their wood early were the most hardy. 

Thayer (70), as a result of observations on winterkilHng of peach 



374 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

buds, found that winter hardiness consists in resistance to the effects 
of warm periods during the winter rather than resistance to low 
temperatures. It is observed that the relative hardiness of varieties 
for a winter continuously cold may be extremely different than for 
a more favorable season. 

The effect of sudden freezes following a warm period is not un- 
known among cereal growers. Georgeson et al. (19) note, for ex- 
ample, that in 1894 the wheat crop at Manhattan, Kans., and else- 
where in the state was almost a failure because of a week of cold 
weather the latter part of March following a period of unusual mild- 
ness. The interesting fact is noted that Turkey wheat survived 
better than others because it was somewhat later in starting growth. 

Nelson (46) attributed the severe injury to winter oats at the 
Arkansas station in 1910-11 to a sudden drop in temperature follow- 
ing a period of good growing weather. 

La Tourette (33) found in experiments previously noted that un- 
germinated wheat was injured much more by freezing if it had been 
soaked in warm water than if it had been soaked in cold water 
previous to freezing. Fleming and the writer (61) froze plants of 
winter wheat, winter oats, winter rye, and winter emmer at a tem- 
perature of — 4 to — 5° C, the plants having been grown in the 
greenhouse. Three separate experiments were performed. In gen- 
eral those grains known to be hardy when grown in the field were 
injured practically the same as the least hardy. 

AGE OF TISSUE. 

Chandler (11) found no constant relation between maturity of 
tissue and resistance to cold. Young leaves of fruit trees were found 
to kill at a higher temperature than old leaves. He also found no 
relation between rate of growth and resistance to low temperature, 
but exposure to low temperature previous to freezing increased the 
resistance. 

Observations on peach buds by Thayer (70) seemed to show that 
buds on young trees are more apt to be injured than on mature trees. 
The well known fact that late-sown fall grains are much less likely 
to survive than those sown at the proper time is probably due mainly 
to the more tender tissue of the young plants. 

CONCENTRATION OF SAP. 

Ohlweiler (51) extracted the sap of a number of trees and shrubs 
and determined the freezing point. From the data collected and 
observations on the injury sustained from a late spring frost he 



salmon: why ckrkai.s wintkkkili., 



arrived at tlu' following' conclusions. ( i ) 'I'liat extreme differences 
in sap density in general are accompanied by a corres])on(ling differ- 
ence in resistance to freezing. (2) That exceptions to this general 
[ rule are probably due to differences in cell structure and other causes 
such as protective locations, etc. (3) That when the cell structure 
is the same the densities of the cell sap indicates the relative hardi- 
ness. (4) That in plants of the same genus or in varieties of the 
same species differences in sap density correspond to differences in 
their resistance to freezing. 

Harris and Popenoe (26) state that sap extracted from the West 
Indian type of Per sea americana froze at a higher temperature than 
that of the Mexican and Guatemalan types, and that horticultural 
experience shows the former to have the least capacity to withstand 
low temperatures. The authors conclude, however, that the cryo- 
scopic constants of the sap does not always determine the degree of 
hardiness. 

Chandler (11) found from extensive experiments with various 
plants that increasing the sap density of easily killed tissue by ab- 
sorption of glycerin, sugars, and mineral salts reduced the injury from 
freezing. 

Wright (79) concluded from experiments with the leaves of wheat 
that the sap froze much more readily when extracted than when con- 
tained in the leaves and that the power to resist freezing is increased 
by exposure to low temperatures. 

Fleming and the writer (61) found that if plants of wheat grown 
in the greenhouse were frozen when turgid they were killed at a 
temperature only a few degrees below zero, but if they were allowed 
to become wilted before freezing they were injured very slightly or 
not at all. Comparisons of the freezing point of the extracted sap of 
different kinds of winter grains failed to indicate any significant rela- 
tion to winter resistance. 

STRUCTURE OF TISSUE. 

Wiegand (74) studied the formation of ice in cell tissue under 
the microscope and determined the size of the cells and the moisture 
content of the tissue. Of twenty-seven trees examined there were 
only eight in which no ice was found in the buds at — 18° C. In 
four of the eight, minute ice crystals were found at — 26.5° C. A 
comparison of seven of those in which ice was formed at — 18° C, 
and of seven in which no ice was found at that temperature showed 
the latter to contain much less water and to have thicker cell walls 
and slightly smaller cells. 



3/6 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Beach and Allen (5) found a correlation between the density of the 
wood of apple trees and hardiness, but exceptions were found. The 
hardier varieties were found to evaporate water less rapidly than the 
less hardy sorts and hence were better able during cold weather to 
maintain a balance between absorption and transpiration. This dif- 
ference was thought, however, more likely to be due to a greater 
concentration of the sap than to a difference in structure of the 
wood. 

MacFarlane (38) observes that "all thermo-resistant plant struc- 
tures are said to have a rich and relatively dense protoplasm or n 
stored mass of reserve material in the cells that contribute to their 
thermo-resistant qualities. These qualities are aided by the occur- 
rence of mucilaginous walls or cell contents, thick, and pigmented 
cellulose or cuticularized walls." 



Conclusions regarding the causes of winterkilling of cereal crops 
at this time would be decidedly premature. There can be scarcely 
any doubt that death occurs as a result of heaving of the soil, 
smothering, and direct effect of low temperature on the protoplasm. 
No doubt physiological drought causes injury and differences in 
resistance of certain cereals may perhaps be explained by their ability 
to absorb a larger quantity of water from the soil in proportion to 
the amount transpired. 

The duration and intensity of cold, rate of freezing, and in certain 
cases the rate of thawing and protection afforded by mulches, snow, 
and uneven surface of the gound are important factors. The mois- 
ture content of the tissue and its condition with respect to dormancy 
often have a determining influence. 

The following outline indicates the probable relation of the dif- 
ferent factors. 



CONCLUSIONS. 



Heaving 
Smothering 



'Desiccation 
Chemical effect of 



Causes of winterkilling ■< 



cold 

Metabolism at low 



Direct effect of low temperature 
Physiological drought 
'Duration and intensity of cold 
Rate of freezing 
Rate of thawing 
Protection 

Kind of soil and moisture con- 



temperature 



Conditions which modify 
the degree of injury.. . 



tent of soil 
Habit of growth of plants 
Moisture content of tissue 
Dormancy 
Age of plants 
Concentration of sap 
Structure of tissue 



salmon: WHY ci'.kiiai.s winterkill. 377 

I.ri l K.M IIUK ClTKI). 

1. :\n\w., C Inlluoncc of cdUI on plant's — a resume. In U. S. Dept. A^r., K.xpt. 

Sla. Record, 6: 777-7S2. i8<)5. 

2. Ad.'Vm.s, J. The effect of hnv temperatures on moist seeds. In Sci. Proc. 

Roy. Dublin Soc., n. s., i.i (1905), no. i, pp. 6. Abs. in Expt. Sta. Record, 
17: 653. igo6. 

3. Allen, F. W. Factors correlated with hardiness in the api)le. In Proc. 

Soc. Hort. Sci., 11: 130-137. 1914. 

4. Babcock, F. R. Cereal experiments at the Willison substation. U. S. Dept. 

Agr. Bui. 270, p. 23. 1915. 

5. Beach, S. A., and Allen, F. W.. jr. Hardiness in the apple as correlated 

with structure and composition. Iowa Agr. Expt. Sta. Research Bui. 21, 
p. 185. 1915. 

6. Becquerel, p. The action of liquid air on the life of seeds. In Compt. 

Rend. Acad. Sci. (Paris), 140 (1905), no. 25, p. 1652-1654. Abs. in Expt. 
Sta. Record, 17 : 653. 1906. 

7. BouYOUCOS, G. J. Soil temperature. Mich. Agr. Expt. Sta. Technical Bui. 

26, p. 132. 1916. 

8. Blackman, F. F. Vegetation and frost. In New Phytologist, v. 8, nos. 

9^10, p. 354. 1909. 

9. Carleton, M. a. The small grains, p. 76. The Macmillan Company, New 

York. 1916. 

10. . Op. cit, p. 298. 

11. Chandler, W. H. The killing of plant tissue by low temperatures. Mo. 

Agr. Expt. Sta. Research Bui. 8. 1913. 

12. Clark, J. A. Cereal experiments at Dickinson, N. Dak. U. S. Dept. Agr. 

Bui. 33, p. 24. 1914. 

13. Delwiche, E. J., and Moore, J. G. The relation of orchard cover crops to 

soil moisture and soil freezing. In 24t'h Ann. Kept. Wis. Agr. Expt. 
Sta., p. 385. 1907. 

14. DE Candolle, D. Arch, des Sci. Phys. et Nat., vol. 2, p. 629 (1879) ; vol. 

33. P- 479 (1895). Cited by Chandler (11). 

15. Detmer, W. Forsch. Geb. Agr. Phys., 9: 292. 1896. Cited by Abbe (i). 

16. Federoff, D. D. Khoziaistvo, no. 12. 191 1. 

17. Garcia, Fabian, and Rigney, J. W. Hardiness of fruit buds and flowers 

to frost. New Mex. Agr. Expt. Sta. Bui. 89. 1914. 

18. Gassner, G., and Grimme, C. The study of cold resistance by cereals. In 

Ber. Deut. Bot. Gesell., 31 (1913), no. 8, p. 507-516. Abs. in Expt. Sta. 
Record, 30 : 524. 1914. 

19. Georgeson, C. C, Burtis, F. C, and Otis, D. H. Experiments with wheat. 

Kans. Agr. Expt. Sta. Bui. 47, p. 16. 1894. 

20. Geoppert, . tiber die Warmeentwickelung in dem Pflanzen ; deren Ge- 

frieren und die Schutzmittel gegen dasselbe. 1830. Cited by Chandler 
(II). 

21. GoRKE, H. tiber Chemische vorgane beim Erfrieren der Pflanzen. In 

Landw. Versuchs., 65: 149. 1906. Cited by Chandler (11). 

22. Green, W. J., and Ballou, F. H. Winterkilling of peach trees. Ohio Agr. 

Expt. Sta. Bui. 157. 1904. 

23. Haberland, G. Physiological plant anatomy, p. loi. Trans, from fourth 

German ed. by Montagu Drummond. Macmillan & Co., London. 1914. 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

24. . Op. cit., p. 138. 

25. Hall, A. D. The soil, an introduction to the scientific study of the growth 

of crops, p. 124. J. Murray, London. 1908. 

26. Harris, J. A., and Popenoe, W. Freezing point lowering of the leaf sap 

of horticultural types of Persea americana. In U. S. Dept. Agr., Jour. 
Agr. Research, 7: 261-268. 1916. 

27. Hickman, J. F. Experiments in wheat seeding, including treatment of 

seed for smut. Ohio Agr. Expt. Sta. Bui., 2d ser., vol. 4, no. 4 (Bui. 
30), P- 77-89- 1891- Field experiments with wheat. Ohio Agr. Expt. 
Sta. Bui. 82. 1897. 

28. HiLLiARD, C. M., ToRAssiAN, C, and Stone, Ruth P. Notes on the factors 

involved in the germicidal effect of freezing and low temperatures. In 
Science, n. s., v. 42, no. 1091, p. 770, 771. 1915. 

29. Hume, A. N., Champlin, Manley, and Morrison, J. D. Winter grains in 

South Dakota. S. Dak. Agr. Expt. Sta. Bui. 161, p. 250. 1915. 

30. Hunt, T. F. The cereals in America, p. 71. Orange Judd Co., New York. 

1908. 

31. JosT, L. Plant physiology. English translation by Gibson, p. 301. The 

Clarendon Press, Oxford. 1907. 

32. KoLKUNov, V. On the selection of a type of wheat resistant to severe 

winters. In Khoziaistvo, v. 7, no. 36, p. 1161-1167; ^hs. in Internat. Inst 
Agr. (Rome) Bui. Bur. Agr. Intel, and Plant DjJeases, 3, no. 12, p. 2631^ 

2634. I9I2. V 

33. La Tourette, L. D. Some factors influencing the resistance of wheat to 

low temperatures. Thesis for M.S. degree, Kans. State Agr. College. 
1915. 

34. LiDFORSS, B. Die wintergriine Flora. In Lund's Universitats Arsskrift, 

Bd. II, no. 13. 1907. Ahs. by Blackman (8). 

35. Lyon, T. L., Fippin, E. O., and Buckman, H. O. Soils, their properties 

and management, p. 306. The Macmillan Co., New York. 1915. 

36. McClelland, C. K. Winter crops. Ga. Agr. Expt. Sta. Bui. 117, p. 337. 

1915. 

37. McFayden, a., and Rowland, S., On the influence of the prolonged action 

of the temperature of liquid air on microorganisms and on the effect of 
mechanical trituration at the temperature of liquid air on photogenic 
bacteria. In Proc. Roy. Soc. London, 71 : 76, 77. 1902. Also in Nature, 
63 : 3. 1900-01. 

38. Macfarlane, J. M. The relation of plant protoplasm to its environment. 

In Jour. Acad. Nat. Sci. Phila., 2d ser., 15 : 249-271. 1912. Abs. in 
Expt. Sta. Record, 28: 326-327. 1913. 

39. Maximow, N. a. Experimental and critical studies on freezing and frost 

killing in plants. In Jahrb. Wiss. Bot. (Pringsheim) 53, no. 3, p. 327- 
420. 1914. Abs. in Expt. Sta. Record, 31 : 34. 1914. 

40. . The chemical protection of plants against freezing. In Russ. Jour. 

Expt. Landw., v. 13, no. i, p. 1-26; no. 4, p. 497-525. Abs. in Ber. Deut. 
Bot. Gesell., 30, no. 2, p. 52-65 ; no. 6, p. 293-305 ; no. 8, p. 504-516. 1912. 
Also abs. in Expt. Sta. Record, 27:333 (1912) ; 28:330-331; 28:630-631 
(1913)- 

41. Mez, C. Neue Unt'ersuchengen iiber das Erfrieren eisbestandiger Pflanzen. 

In Flora, 94: 89. 1905. Cited by Chandler (11). 



salmon: why ci:ki;als vvi N'ri:in<ii-L. 



379 



4-'. Moi.iscii, . UntcrsuoluiiiK iilur das I'j-f ric-rcn dcr l^flanzcn. 1H07. 

Citi'd hy Chandler (11). 

43. MoNTGOMKRV. K. (i. Rolliii.i; wiiitcT wheal. Nehr. A^r. ICxpt. Sta. I'rcss 

Bui. 30. igoy. 

44. . Productive farm crops, p. uX. |. 1',. Lii)pinc()tt Co., Phila. i(>i6. 

45. Mvller-Thukgau, H. Uher das Ciefrieren und Erfricrcn der JMIanzen. 

In Laiulw. Jahrb., 9: 133 (1880); 15: 4S3 (1886). Cited by Chandler 
(II). 

46. Nelson, Martin. Oats. Ark. A^^r. Expt. Sta. Bui. 118. 1914. 

47. Newman, J. S., and Pickett, J. S. Wheat. S. C. Agr. Expt. Sta. Bui. 56, 

p. 5. 1900. 

48. Neveroff, a. J. Experiments in catching and holding snow on fields in 

southwest Russia. Zeml. Gaz., no. 46. 1914. 

49. Noll, C. F. Tests of wheat varieties. In Ann. Rpt. Pa. State Coll. and 

Agr. Expt. Sta. (1912-13), pt. 2, p. 47-55- IQM- 

50. OsKAMP, Joseph. Soil temperature as influenced by cultural methods. In 

U. S. Dept. Agr., Jour. Agr. Research, 5: 173. 1915. 

51. Ohlweiler, W. W. The relation between density of cell sap and the freez- 

ing point of leaves. In 23d Ann. Rept. Mo. Bot Gardens, p. 101-131. 
1912. 

52. Patten, H. E. Heat transference in soils. U. S. Dept. Agr., Bur. Soils 

Bui. 59, p. 10. 1909. 

53. Pictet, R. Archives des Sci. Phys. et Nat., 11: 320. 1884. Also in ser. 

3, V. 30, p. 293. 1883. Cited by Schimper (65). 

54. Piper, C. V. Forage plants and their culture, p. 374-412. The Macmillan 

Co., New York. 1915. 

55. Pfeffer, W. Physiology of plants, v. 2, p. 221. Trans, by E. J. Ewart. 

Oxford, 1903. 

56. Sachs, J. Krystallbildungen bei dem Gefrieren und Veranderung der Zell- 

haute bei dem Aufthauen saftige Pflanzen-thiele. Ber. Ver. Kon. Sachs. 
Gesell. Wiss. zu Leipsig, 12: 1-50. i860. Also Landw. Versuchs., v. 5, 
p. 167. i860. Cited by Chandler (11). 

57. Salmon, S. C. Seeding winter grain in furrows to prevent winterkilling. 

In Jour. Amer. Soc. Agron., 8: 176. 1916. 

58. . The relation of soil type and moisture content to temperature and 

winterkilling. In Science, n. s., (in press). 

59. . The relation of winter temperature to the distribution of winter and 

spring wheat. In Jour. Amer. Soc. Agron., 9: 21-24. 1917. 

60. . Unpublished data in Office of Cereal Investigations, U. S. Dept. Agr. 

61. . Unpublished data, Kans. Agr. Expt. Sta. 

62. Schaffnit, E. tiber den Einfluss neider Temperaturen auf die Pflanzen. 

Zell. Zeits. Allg. Phys., 12: 323-336. 1912. Cited by Chandler (11). 

63. . In Jahresber. Kaiser Wilhelms Inst. Landw. Bromberg, p. 21-23. 

1913. Abs. m Expt. Sta. Record. 33: 51. 1915. 

64. . tiber den Einfluss nieder Temperaturen auf die Pflanzliche Zelle. 

Mitt. Kaiser Wilhelm Inst. Landw. Bromberg, Bd. 3, Ht. 2, p. 93-144. 
Abs. in Internat. Inst. Agr., Bur. Agr. Intell. and PI. Diseases, 2: 47. 
Also abs. in Expt. Sta. Record, 24: 533. 191 1. 

65. Schimper, A. F. W. Plant geography upon a physiological basis, p. 25-41. 

Trans, by W. R. Fisher. The Clarendon Press, Oxford. 1903. 



380 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



66. Selby, a. D. Fall and winter injury to orchard trees and shrubbery by 

freezing. Ohio Agr. Expt. Sta. Bui. 192, p. 129^-148. 1908. 

67. Shutt, F. T. On the relation of moisture content to hardiness in apple 

twigs. In Proc. and Trans. Roy. Soc. Canada, ser. 2, v. 9, Trans, sec. 4, 
p. 149-153. 1903- 

68. SiNZ, E. Bezichungen zwischen Brocksubstang und Winter Festigkeit bei 

verschiedewen winterweizen Varietaten. In Jour. Landw., 62: 302. 1914. 
Ahs. in Expt. Sta. Record, 33 : 235. 1915. 

69. Smith, J. W. Agricultural meteorology. U. S. Dept. Agr., Weather Re- 

view, 44, no. 2, p. 1475. 1916. Ahs. in Expt. Sta. Record, 35: 114. 1916. 

70. Thayer, P. Winterkilling of peach buds. Ohio Agr. Expt. Sta. Mo. Bui. 

I, no. 10, p. 311, 312. 1916. 

71. Waldron, L. R. In N. Dak. Agr. Expt. Sta. Ann. Rpt'. Dickinson (N. 

Dak.) substation for 1910, p. 32. 191 1. 

72. Watson, G. C, and Miller, N. G. Variety tests of wheat, oats and pota- 

toes. Pa. Agr. Expt. Sta. Bui. 76, p. 4. 1906. 

73. WiEGAND, K. M. Some studies regarding the biology of buds and twigs in 

winter. In Bot. Gaz., 41 : 373-424. 1906. 

74. . The occurrence of ice in plant tissues. In Plant World, 9 : 31-32. 

1906. 

75. WiESNER and Packer. Ueber die Transpiration entlauber Zweige und des 
. Stammes der Rosskastanie. In Oesterr. Bot. Zietschr., no. 5. 1875. 

Cited by Haberlandt (23), p. 138. 

76. Wiggans, C. C. Studies regarding the rest period of seeds. Thesis, Univ. 

of Mo., 1913. Cited by Howard, Mo. Agr. Expt. Sta. Research Bui. 21, 
p. 42. 1915. 

77. Winkler, A. Influence of external conditions on the resistance to cold of 

perennial plants. In Jahrb. Wiss. Bot. (Pringsheim), 52, no. 4, p. 467- 
506. Abs. in Expt. Sta. Record, 30: 333. 1914. 

78. WoLLNY, C. Forsch. Geb. Agr. Physics, 9: 292. 1886. Cited by Abbe (i). 

79. Wright, S. G. Relation of low temperature to the growth of wheat. In 

Agr. Sci., 4: 337. 1890. 



AN ANNUAL VARIETY OF MELILOTUS ALBA.^ 

H. S. CoE. 

In the winter of 1916 a quantity of Melilotus alba seed which had 
been grown in Hale County, Ala., the previous summer was purchased 
for experimental purposes. This, together with seed which had been 
grown in Mississippi, Kentucky, Kansas, Montana, Wyoming, South 
Dakota, and North Dakota, was sown in adjacent plats at Redfield, 

1 Contribution from the Office of Forage-Crop Investigations, Bureau of 
Plant Industry, U. S. Department of Agriculture, Washington, D. C. Received 
for publication October 22, 1917. 



COK: an ANNUM- W lil l l-; SWKKT C LOVKU. 3<S I 

S. Dak., on Ai)ril 15, i«)if>. and al l-'ar^o, N. Dak., on May 27, 
No nnrsc crop was used at cillicr place. 

Al RiHllicld. ai)])roNiniaU'ly 5 jxTrcnt of tlic ])Jaiils on llu- plat 
sown with .Mabania seed llowcrcd al)nndantl\ and niatnrcd seed in 
Sci)tcMnl)cM-, i()i(). riio remainder ol" the plants on tins i)lat, as well 
as all the i)lants on the other seven plats, made a tyi)ical hrst year's 
growth for Mclilotiis alba. Dnrini;- the first year typical Mclilotus 
alba i)lants prodnce an npright, leafy, branching growth bnt no 
llowers. The taproot becomes much enlarged at the crown and in 
late summer or early autumn a varying number of buds form at the 
crown and serve to produce the first growth the following season. 
The plants which flowered grew to a height of 3.5 to 4 feet, whereas 
those which did not bloom made an average growth of 2 to 2.5 feet. 

At Fargo, the same as at Redfield, approximately 5 percent of 
the plants on the plat sown with Alabama seed flowered abundantly. 
By August 21 the plants on all of the plats had made a 36-inch 
grow'th and were cut for hay, leaving an ii-inch stubble. From that 
time until October 4 the annual plants made a second growth of 18 
to 24 inches, whereas the biennial plants made no more than a 6-inch 
growth. The second crop of the annual plants bloomed profusely 
but the pods did not mature on account of frost. All plants in the 
other seven plats made a typical first year's growth ior Melilotus alba. 

A careful examination of the plants which flowered at both sta- 
tions showed that in most respects their botanical characters were in- 
distinguishable from those of a second year's growth of Melilotus 
alba. The leaves of the plants which flowered the first year under 
field conditions, as well as the leaves of the plants grown in the green- 
house the following winter from the seed of these plants, were as a 
whole more oblanceolate and more distinctly serrated than the leaves 
of the second year's growth of the typical Melilotus alba plant. 
There was no difference in the venation of the seed pods, a char- 
acter which is used often to distinguish the different species of Meli- 
lotus. However, the most striking difference between the plants 
which flowered the year of seeding and those which did not flower 
until the summer of 1917 was in the type of root produced. Without 
a single exception the plants which bloomed the first year at both 
Redfield and Fargo produced typical annual taproots with no en- 
largements at the crowns and with no crown buds. None of these 
plants lived through the winter of 1916-17, whereas only a small 
percentage of the normal plants of Melilotus alba winterkilled. 

Seed was collected from a number of the annual plants at Redfield 
in October, 1916. On January 27, 1917, 275 seeds were sown in in- 



382 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



dividual pots in the Department's greenhouse. Plants were produced 
in 255 pots and by May 25 they had made a growth of 4.5 to 5.5 feet 
and were in bloom. Examination of the roots of these plants showed 
that they wefe typical annual taproots and that the root growth com- 
pared favorably with that made by the parent plants under field con- 
ditions. An attempt has been made at different times to have typical 
Melilotus alba bloom directly from seed in the greenhouse but the only 
flowers that have been obtained thus far have been produced from 
branches from crown buds after the plants had been permitted to 
pass through a resting stage. Under field conditions the only flowers 
which have been obtained on the first year's growth of Melilotus alba 
were on plants which were started in the greenhouse early in Feb- 
ruary, 1916, and which were transplanted to the nursery at Arlington 
Farm in April. Three of these plants out of a total of 100 produced 
several racemes apiece in October, but they also produced an abun- 
dance of crown buds and lived through the winter. 

It is very possible that the white-flowered annual variety may be 
found in more localities of the South than the one mentioned, as sev- 
eral letters and specimens were received the past year from persons 
who had purchased southern Melilotus alba seed and who stated that 
some of the plants bloomed the year the seed was sown. 

It is probable that this plant will be of considerable economic value 
in the southern portion of the Gulf States as a winter legume and in 
the central and northern portions of the country as a summer hay crop 
and for green manure. The acreage of Melilotus alba seeded on 
winter grain or with spring grain to be turned under in the autumn 
for green manure or cut for hay is rapidly increasing. The most 
serious objection to this practice is the difficulty of completely eradi- 
cating the plant by fall plowing. In addition to making more growth 
after harvest than spring sowings of the biennial variety, no trouble 
will be experienced in eradication when the annual variety is used. 

In view of these facts it is believed that this plant is worthy of 
botanical designation. Specimens have been placed in the herbarium 
of the New York Botanic Garden, the National Herbarium, and the 
Asa Gray Herbarium. The following is a description of the plant: 

Melilotus alba Desr. var. annua n. var. (Annual White Sweet Clover). — 
Erect or ascending, branching, glabrous or young branches and leaves slightly 
pubescent; leaves petioled, leaflets mostly oblanceolate, some narrowly ovate to 
oblong, serrated, obtuse to truncate ; corolla white, 4 to 5 mm. long, the stand- 
ard longer than the other petals; racemes numerous, slender, 4 to 15 cm. long; 
pods reticulate, 3 to 4 mm. long; root becoming 15 to 30 inches in length and 
enlarged very slightly if at all at the crown. Crown buds are not formed. 



ACKONOM iC AI'I AI US. 



3«3 



AGRONOMIC AFFAIRS. 

MEMBERSHIP CHANGES. 

The membership reported in the last number was 653. The deaths 
of two members have been reported since that time, and two new 
ones have been added, so that there is no change in the total mem- 
bership. The names and addresses of the new members, the names 
of the deceased members, and such changes of address as have been 
reported to the Secretary follow. 

New Members. 

Jacobs: Daniel C, 112 W. Fairmont Ave., State College, Pa. 
Lund, Viggo, Farm Crops Dept., Iowa State College, Ames, Iowa. 

Members Deceased. 

loughridge, r. h. 
ScHULz, Arthur W. 

Changes of Address. 

Bailey, C. H., University Farm, St. Paul, Minn. 
Derr, H. B., Agricultural Advisor, Fairfax, Va. 
Kemp, W. B., Sparks, Md. 

Parker, John H., Experiment Station, Manhattan, Kans. 

Petry, Edward J., 115 University St., West La Fayette, Ind. 

ScHooNovER, Warren R., c-o A. R. Gawthrop, R. R. i, Milford, Ind. 

Shoesmith, V. M., The Jennings Farms, Bailey, Mich. 

TowLE, R. S.,' Northern Great Plains Field Sta., Mandan, N. Dak. 

Walster, H. L., 5616 Kenwood Ave., Chicago, 111. 

Whiting, Albert L., 705 Gregory St., Urbana, 111. 

WiGGANS, Roy G., Farm Crops Dept., Cornell University, Ithaca, N. Y. 

Woods, Albert F., Maryland State College, College Park, Md. 

NOTES AND NEWS. 

The item regarding W. A. Albrecht in the October Journal 
was in error. Mr. Albrecht still retains his position in the soils de- 
partment of the University of Missouri. 

R. H. Loughridge, a charter member of this Society and professor 
of agricultural chemistry in the University of California from 1891 to 
1909, since which time he has been professor emeritus, died at the 



384 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



home of his brother at Waco, Texas, on July i. Professor Lough- 
ridge -was born at Koweta, near Muscogee, in what was then Indian 
Territory, on October 9, 1843, graduated ^T^om the University of Mis- 
sissippi in 1871, and received the degree of Ph.D. from that institu- 
tion in 1876. He held various positions in geological and chemical 
work in the South from 1871 to 1891, when he was called to the 
University of California by his former teacher and colleague. Prof. 
E. W. Hilgard. Professor Loughridge's work in Cahfornia was 
mostly along the line of soil physics and soil chemistry, particularly 
studies of the arid and alkali soils of California. 

R. A. Gortner, associate professor of agricultural biochemistry in 
the Minnesota college, has been made professor and chief of the 
division of agricultural chemistry in the college and station, succeeding 
R. W. Thatcher. 

John H. Parker, for the past several years assistant in the office of 
cereal investigations, U. S. Department of Agriculture, since Novem- 
ber I has been assistant in cereal breeding at the Kansas college and 
station. 

Warren R. Schoonover, instructor in soil biology in the University 
of Illinois, has enlisted in the gas defense service of the sanitary 
corps, U. S. Army. He will probably be connected with Overseas 
Repair Section No. i. 

Arthur W. Schulz, scientific assistant in dry-land agriculture at the 
Northern Great Plains Field Station, Mandan, N. Dak., was killed by 
a train at his former home in Kansas early in August. 

R. W. Thatcher, for the past several years professor of agricul- 
tural biochemistry in the Minnesota college and vice-director of the 
station, has been elected director and dean of the department of agri- 
culture in the same institution, succeeding A. F. Woods. E. M. Free- 
man, professor of botany and vice-dean, has been elected dean of the 
college of agriculture. 

H. L. Walster, assistant professor of soils in the Wisconsin college 
and station, is on leave of absence and is pursuing graduate study in 
the department of botany of the University of Chicago. 

H. J. Waters, president of the Kansas State Agriculture College 
for the past eight years, has accepted the managing editorship of the 
Kansas City Weekly Star, the weekly agricultural edition of the well- 
known daily of the same name. 



JOURNAL 

OF THE 

American Society of Agronomy 



Vol. 9. December, 1917. No. 9. 



THE AGRONOMIST OF THE FUTURE. 

W. M. Jardine. 

(Presidential Address before the American Society of Agronomy, November 

12, 1917.) 

It is a privilege which I esteem highly to stand today in the pres- 
ence of the assembled representatives of two institutions, the Society 
for the Promotion of Agricultural Science and the American Society 
of Agronomy. For the American Society of Agronomy, which I 
have the honor to represent, it is the tenth annual meeting. The little 
group of earnest men who met in Chicago ten years ago to organize 
this society doubtless had visions of the increased service which 
agronomists might render by such a union of forces. I very much 
doubt, however, whether their imaginations visuaHzed, even as a pos- 
sibility, the nation plunged into a world war in the short space of one 
decade or the service we are now called upon to give. 

Since that fatal 4th of August, 1914, all agencies. Federal, State, 
and private, collective and individual, have been called upon to face a 
great problem. Men must be fed and clothed before they can fight. 
A continuous stream of foodstuffs must be kept moving from this 
country and Canada to our allies and the allied armies at a time when 
not only is the world's available food supply low, but the stores of 
wheat in Russia, India, and Argentina are inaccessible. Especially 
heavy, therefore, is the responsibility resting upon American agri- 
culture. Its problem is not merely one of planting greater acreages 
of food crops, but of increasing the output with a reduced force of 
workmen. With the outbreak of war in 1914, the period of abundant 
labor came to an end. Great war contracts were awarded our manu- 
facturers and available labor was rapidly drawn to industrial centers 

385 



386 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



by the lure of high wages. Canada immediately began sending her 
ablest bodied men across the ocean— 33,000 of them within six weeks 
after war was declared and between 350,000 and 400,000 men by 
April, 1917. Canadian authorities were forced into aggressive action 
to secure men to keep their industries in operation and many men 
crossed over from the States to fill the ranks of workers in Canada. 

Upon our entry into the war, the government had need of the 
immediate services of every industry, every organization, and every 
individual, and the response came promptly and loyally from every 
quarter. Agronomists at once evinced an all-pervading desire to 
bring to the immediate assistance of the country every ounce of their 
strength which might be put to practical use. They frequently proved 
to be the logical men to form " ways and means " committees for 
devising plans whereby the maximum production could be secured 
from every man, every horse, every machine, and every acre of 
ground, quickly, and at the same time jeopardize in no way the per- 
manence of agriculture. Campaigns for increased crop acreages were 
initiated and carried to successful completion. Seed stocks were in- 
ventoried and made available for those in need. Information on 
every conceivable point relating to agriculture was given to the public 
through correspondence, public addresses, and the press. This suc- 
cessful participation of agronomists in emergency work is a splendid 
tribute to their usefulness. It will be continued with increasing vigor. 

In our preoccupation in the present great emergency, however, we 
must not forget the future. During the present crisis and after it has 
passed there are two fundamental ideals which agronomists must 
keep before them. The first is to render such service as will warrant 
the increasing confidence of farmers. The second is to secure so 
thorough a training as to be able to understand and appreciate the 
relationship between the fundamental sciences and agronomy, and to 
make original contributions to the science of agriculture. 

While agronomists have demonstrated their grasp of practical 
affairs in the present crisis, the fact that they have not always been 
taken into the wholehearted confidence of farmers must be recognized. 
College men are still regarded as theorists by a great many farmers. 
It is a mistake to blame this attitude entirely on the farmer. There 
is no doubt that impractical methods have sometimes been advo- 
cated. We have at times dispensed unsound information and every 
foolish statement has offset ten true ones. The true statements are 
accepted by farmers as a matter of course, while the faulty one is too 
often remembered and regarded as an index to the practical intel- 
ligence of the college man. 



jAKiJiNE: TJiic ACiuuNoM 1ST OF riii: M ri'Ki:. 387 

We need not look far for conspicuous examples of such unsound 
advice and impractical methods. For years we have advocated the 
improvement of small grains by mass selection. Until less than live 
years ago there was perhaps no piece of advice so commonly given 
out. Yet it is doubtful if this method is actually followed by one 
tenth of I ])orcent of the farmers of the United States, simply be- 
cause it is impractical, if it is not altogether unsound. What folly 
to advise the farmer who has 500 acres of wheat to harvest, and 
who is employing from eight to ten men and tw^ice as many horses, to 
select enough of the best heads to thrash a bushel of grain ! This is 
a process requiring at least a week's time in the very busiest and most 
critical season of the year. Furthermore, unless he had given the 
matter special attention, it is altogether likely that he would secure a 
poorer rather than a better grade of grain, since he would naturally 
select the largest and best appearing heads, which would result in 
many instances in a later maturing strain and one inferior both in 
yield and in quality. 

The ear-row test of corn is another example. This method of im- 
provement has been generally advocated, at least until the last few 
years, but farmers are not using it because it will not stand the test 
of practical application and will not fit into the system of farming 
used by farmers. It is a satisfactory system for the man who 
specializes on corn growing or who grows corn for seed, but not for 
the ordinary farmer. 

Then, there is the soil mulch. We have advocated this practice 
strongly and it has been widely accepted and adopted by farmers. 
Now we are not so sure that the soil mulch really conserves moisture 
by preventing evaporation or that it aids in nitrate formation. Its 
value may come mainly from the destruction of weeds. If we must 
admit there is not so much to the soil mulch as we formerly thought, 
it will tend to shake the farmers' confidence in us, as well as in the 
mulch. 

How does it come that we find ourselves in the position of having 
made assertions which we are under the necessity of retracting? The 
answer in the first .two cases mentioned — the improvement of small 
grains by mass selection and the ear-row test of corn — is undoubtedly 
the lack of a thoroughly practical attitude toward the farmers' prob- 
lems. The last case, that of the soil mulch, probably arose largely 
from the lack of that fundamental, scientific training which will not 
allow a problem to be dropped until the why as well as the how has 
been discovered. The agronomist is not alone in having given out 



388 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

faulty information. Chemists, physicists, and other scientific work- 
ers have made similar errors, but this does not excuse the agronomist. 

In common with other agricultural workers the agronomist holds a 
somewhat anomalous position. He is more or less of a middleman 
between the man of pure science and the man on the farm. He must 
be scientific enough to understand the principles evolved by the 
pathologist, the biologist, and the chemist, and practical enough to 
apply these principles to the business of farming. 

A practical point of view is largely the result of boyhood training 
and an endowment of common sense. The colleges may develop a 
practical attitude but they cannot endow the constitutionally imprac- 
tical boy with common sense, brilliant though he may be. As the 
first requisite of a keen cutting edge is hard steel, so the first qualifica- 
tion of the future agronomist is going to be an abundant endowment 
of common sense. One cannot put a good edge on soft steel and we 
who are now directing the training of future agronomists should 
make it our first article of faith to advise a boy to go into teach- 
ing or research in agronomy only when he has shown ample evidence 
of his thorough-going practicality. In the last analysis, of course, 
practical information is that which is scientifically correct. I submit, 
however, that only the acid test of actual farm practice can prove any 
method, any strain of seed, or any mechanical device to be scientifically 
correct so far as the farm is concerned. I hasten to caution you who 
are finding in your experimental studies, new and promising varie- 
ties of crops, or new and promising methods of soil tillage — such, for 
example, as that of seeding wheat in furrows as a means of over- 
coming winterkilling and soil drifting — against preaching their gen- 
eral adoption until their usefulness has been thoroughly demonstrated 
on the farms by the farmers themselves. 

I suggested a moment ago that the reason why we might have to 
reverse ourselves in the matter of the soil mulch was that we had not 
followed the matter clear through and determined why the mulch gave 
certain results. There are other " whys " that are being pressed per- 
sistently upon us. Last year we lost approximately 12,000,000 acres 
of wheat from winterkilling, the fundamental causes of which no one 
can explain satisfactorily. The ef¥ect of soil acidity on plant growth 
is a problem that awaits solution. We have no generally accepted 
method of detecting soil acidity or of determining the lime require- 
ment of soils. What is perhaps the most important problem in soil 
fertility is no nearer solution today than it was seventy-five years 
ago when Liebig advanced the theory that crop adaptation and ferti- 



jakdink: tiik acronomist oi- Tin-: futuue. 389 

lizer requirements of the soil c(;ul(l be delennined by ehemical 
analysis. While investigators since Liebig's time have corrected many 
of the errors which he made and have shown the importance of 
physical texture and the presence in the soil of organic matter in a 
state of decay, yet with respect to many of the soil problems con- 
fronting the farmer, they can only give general directions, concerning 
the why " of which they have only vague knowledge. 

Turning from the soil to crops again, what do we know about why 
one strain of corn outyields another? Perhaps we have gone into the 
matter far enough to say that one strain is more resistant to smut, 
insects or drought than the other, but this only presses the matter 
one step further back and we are again met with' an insistent " why." 
Possibly these are but manifestations of differing hereditary consti- 
tutions and these characteristics are directly dependent on definite 
genetic factors discoverable only by precise analysis. 

These problems w^ill not be solved by men just out of college, with 
only a bachelor's degree. The main responsibility of agricultural 
colleges is, and will continue to be, the training of young men for 
rural leadership, as farmers and teachers, and for these our present 
four-year curriculum is admirably adapted. It provides sufficient 
electives to permit specialization in animal husbandry, or crops and 
soils, as the young men may choose, and, on the whole, fits young men 
to succeed in the practice of diversified farming or in general agri- 
cultural teaching. It does not, however, fit men to solve fundamental 
problems. 

In the past, leading men in agriculture have not always been well 
trained. Because of native ability, practical experience, and the fact 
that the subject was imperfectly developed and had many pressing 
problems of a very elementary character, these men with limited 
training have achieved notable results. The problems of the future 
will be more difficult and the meager training of the past will not 
suffice. Perhaps two thirds of our present agronomists have no train- 
ing beyond that indicated by the bachelor's degree. They labor under 
the handicap of continually being assigned problems for the solution 
of which they are not adequately prepared. 

With a practical attitude towards the farmer's problems and a 
practical agricultural education as a foundation, the agronomist who 
is to achieve notable results in the future must undergo a thorough 
training in the sciences fundamental to agronomic problems. If crops 
are his particular interest he not only will need to have a solid ground 
work in the general sciences but also a specialization in plant breeding, 



390 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

plant physiology, and plant pathology. If his interest turns towards 
soils he will need, in addition to fundamental knowledge, special 
training in chemistry, physics, and bacteriology. 

The responsibility for advanced training rests not alone upon the 
young men who are ambitious to become agronomists but upon us 
who are already in the field. We may well ask. What about the 
supply of young men who will constitute the agronomists of the 
future? To what extent will the war reduce the number of young 
men entering colleges for agricultural training? An inquiry made 
since the opening of the present school year reveals the fact that the 
present enrollment in 21 of the leading agricultural colleges of the 
country is 68 percent of what it was at this time last year. In these 
21 institutions 6,823 young men are enrolled for agricultural work as 
against 10,011 last year at this time. Many young men are at a loss 
to know whether they should continue their college work or enlist 
for military service. The organizations represented here today 
should take an active interest in this situation. As far as practicable, 
young men should be encouraged to continue scientific training in 
agriculture in order to be prepared to help solve the problems of the 
great period of reconstruction which must follow the end of the war. 
The American Society of Agronomy must encourage its own mem- 
bers to pursue advanced study. To this end, it should encourage the 
policy on the part of educational institutions of permitting members 
of the teaching and research staff to spend time in study at other col- 
leges and universities on part salary. The time has come when the 
institution that does not extend this privilege to its members and 
stimulate their acceptance of the privilege will find itself in a state 
of retrogression. 

Never before has the opportunity for service been greater for 
agronomists. The opportunity to do constructive work in one of the 
most critical periods in the history of human progress is one to be 
considered gravely and with renewed courage and determination. 
The world's difficulties will not all pass with the ending of the war. 
The making of " two blades of grass to grow where but one grew 
before " must be accomphshed if a hungry world is to be fed at a cost 
small enough to allow a safe margin for the attainment of happiness. 
This is the agronomist's problem. It is the agronomist's opportunity. 



Ki:i'()RT oi' Till-: si:( ki:tauv 



AGRONOMIC AFFAIRS. 

REPORT OF THE SECRETARY FOR 1917. 

This meeting marks the tenth anniversary of the founding of the American 
Society of Agronomy. While the actual organization was not effected until 
December 31, 1907, the plans were under discussion for several weeks previous 
and officially we are now closing our tenth year. Though conditions through 
the greater part of the year have been unusual and trying, the Society has con- 
tinued to increase in membership, while the interest in the Journal has been 
greater than ever. The war has had its effects on the Society as on everything 
else, in keeping down the number of new members and perhaps in the loss of a 
few old ones, though most of the losses will not be recorded until after the 
beginning of the next year. High prices have also entered into our affairs, 
preventing in part the increase in the size of the Journal, 

The increased membership and the greater frequency of issue of the Journal, 
together with the greater number of papers handled, has again increased the 
work of the Secretary and Editor. That official again recommends the separa- 
tion of the two offices, a recommendation which he will make in detail to the 
Executive Committee, as that body has the power to choose the editor. As in 
previous years, the Secretary wishes to acknowledge the hearty cooperation 
of the Executive Committee and of the Treasurer in particular, and to express 
his thanks to Misses Jane B. Taylor and Elizabeth C. Lambert for efficient 
service rendered. He also wishes to take this occasion to express his gratitude 
to the entire membership for the support they have given him through the 
three years of his incumbency, the demands of which he feels he can no longer 
meet with fairness to himself and to the Society. 

I. FUNDS COLLECTED BY THE SECRETARY 

The following is a classified list of funds which have been received by the 
Secretary, chiefly from dues of new members and from the sale of Proceedings 
and Journal. All these have been transmitted to the Treasurer and have been 
included in his annual report. 

Classified Receipts and Disbursements, October 28, 1916-OcTOBER 31, 1917. 

Receipts. 

To dues collected (itemized list appended) : 

3 new members for 1916 at $2.00 $ 6.00 

106 new members for 1917 at 2.00 212.00 

4 new members for 1917 at 1.50* 6.00 

3 local members for 1916 at .50 1.50 

24 local members for 1917 at .50 12.00 $237.50 

" Members of local sections who had previously paid dues of 50 cents each as 
local members. 



392 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



To Proceedings and Journal sold: 

14 copies of Volume i at $1.00 $ 14.00 

3 copies of Volume i at 1.80^ 5.40 

4 copies of Volume i at 2.00 8.00 

13 copies of Volume 2 at i.oo 13.00 

2 copies of Volume 2 at 1.80* 3.60 

4 copies of Volume 2 at 2.00 8.00 

14 copies of Volume 3 at i.oo 14.00 

3 copies of Volume 3 at 1.80^ 5.40 

4 copies of Volume 3 at 2.00 8.00 

13 copies of Volume 4 at i.oo 13.00 

3 copies of Volume 4 at 1.80^ 5.40 

,4 copies of Volume 4 at 2.00 8.00 

II copies of Volume 5 at i.oo 11.00 

4 copies of Volume 5 at 1.80^ 7.20 

8 copies of Volume 5 at 2.00 16.00 

II copies of Volume 6 at i.oo 11.00 

4 copies of Volume 6 at 1.80^ 7.20 

8 copies of Volume 6 at 2.00 16.00 

10 copies of Volume 7 at i.oo 10.00 

3 copies of Volume 7 at 1.80^ 5.40 

9 copies of Volume 7 at 2.00 18.00 

9 copies of Volume 8 at i.oo 9.00 

4 copies of Volume 8 at 1.80^ 7.20 

8 copies of Volume 8 at 2.00 16.00 

44 copies of Volume 9 at 1.25^ 55-00 

ID copies of Volume 9 at 1.70^^ 17.00 

21 copies of Volume 9 at 1.80^ 37-8o 

56 copies of Volume 9 at 2.00 112.00 

I copy of Volume 10 at 1.80^ 1.80 

3 copies of Volume 10 at 2.00 6.00 

13 single numbers of Volume 5 to 9 4.40 $473.80 

To extra reprints sold: 

From January-February, 1916, number 5.70 

From March-April, 1916, number 13.53 

From May-June, 1916, number 4.41 

From July-August, 1916, number 1.78 

From September-October, 1916, number 2.53 

From November-December, 1916, number 1.93 

From January, 1917, number 1.18 

From February, 1917, number 1.00 

From March, 1917, number 3.02 

From April, 1917, number 17.80 

From May, 1917, number 12.11 

From October, 1917, number 2.10 $ 67.09 

To sales of copper and zinc from old engravings 1.98 

$780.37 

* Sold through agents at 10 percent discount. 

^ Sold to senior students in agronomy in accordance with decision of Execu- 
tive Committee. ^ Sold through foreign agents at 15 percent discount. 



Kicroirr oi- imii". siccrktauy. 393 
I )ishi(rsi'i)u'itts. 

Jan. 2, 1917, by check to Treasurer Roberts $I59-3I 

Feb. 28, 1917, by check to Treasurer Roberts 307.76 

May 10, 1917, by clieck to Treasurer Roberts 110.60 

June 30, 1917, by check to Treasurer Roberts 72.36 

Oct. 31, 1917, by check to Treasurer Roberts 130.34 $780.37 

Balance on hand Novenil)er i, 191 7 $000.00 



2. MEETINGS. 

Only one meeting of the Society has been held this year, the present annual 
meeting. There have been 22 papers presented, 19 in separate sessions of this 
Society and 3 in a joint session with the Society for the Promotion of Agri- 
cultural Science. That the Society is fully alive to present-day problems and 
particularly problems which have to do with the food supply is shown by the 
unusual preponderance of papers having to do with some phase of wheat pro- 
duction. The average attendance at the sessions of this Society has been about 
90, while the attendance at the joint session was about 225. 

3. LOCAL SECTIONS. 

The Society now has 11 local sections, 2 new ones having been formed during 
the year, one at the University of Illinois and one at the North Carolina A. 
and M. College. The local sections of the Society are now located at Cornell 
University, Georgia State College, the University of Illinois, Iowa State Col- 
lege, Kansas State Agricultural College, Minnesota College of Agriculture, 
North Carolina A. and M. College, Ohio State University, South Dakota State 
College, in New England, and in Washington, D. C. These local sections report 
memberships ranging from 10 to about 65. Most of them hold several meet- 
ings a year, at which problems of local agronomic interest are discussed. The 
membership at several other institutions is sufficient for the formation of local 
sections, and it is hoped that the coming year will witness still further growth 
along this line. 

4. MEMBERSHIP. 

The increase in membership of the Society was not so great as in 1916, the 
banner year, but is greater than in any other year. The number of new mem- 
bers added was 118, 29 less than in 1916, but 11 more than in 1915 and 47 more 
than in any year previous to 1915. The total number of members added since 
1908, when the Society had a membership of 121, is 707, making a gross mem- 
bership of 828. Of this number, 6 have died, 49 have resigned, and 121 have 
allowed their membership to lapse, a net loss of 176 and a present membership 
of 652. The unusually heavy loss by resignation during the year, 24 members, 
was due to a combination of causes, but perhaps more than anything else to 
the shifting of men into emergency agricultural demonstration or military 
work. 

In this connection, it is suggested that the Society make some provision for 
carrying those of its members engaged in military, service on its rolls during 
the period of the war without cost to them. If these men return to agronomic 
work they should be allowed to renew their memberships without payment of 
arrearages and, if they desire, to purchase volumes published during their 
absence at the usual reduced price to members. 



394 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



No special campaign for members was made during the year, other than the 
sending out of sample copies and circular letters. The increase in membership 
is due to these, to the efforts of the individual members, and to the further 
growth of the local sections. A few members have been particularly diligent 
in adding to our rolls, and grateful acknowledgment is here made to them. In 
almost every agricultural college and experiment station in the United States 
and in Canada men are engaged in agronomic work who are not members of 
this Society, not to mention hundreds in private work. The membership of 
the Society should soon reach the thousand mark, but it will not reach that 
mark soon unless the individual members make special efforts to that end. 



Of the io8 changes of address recorded in the Journal during 1917, 83 were 
due to change of position. This number is 13 percent of the total membership, 
as against 11 percent of similar changes in each of the two years previous, 
showing rather more than the usual activity among agronomists. 



The total membership at the end of 1916 was 586. During 1917, 118 new 
members have joined the Society and 2 previously lapsed have been reinstated. 
During the same period 3 have died, 40 have lapsed (of whom 13 were later, 
reinstated), and 24 have resigned. The total loss during the year is 54 and 
the net gain 66, making a total membership of 652. A list of new members for 
1917 follows. The addresses of all new members will be found in the various 
issues of the Journal during the year. Those marked with an asterisk are 
reported for the first time in this issue. 



Changed Addresses. 



New Members. 



New Members in 1917. 



Agee, John H. 
Albert, A. R. 
Anderson, A. C. 
Andrews, Myron E. 
Baker, O. E. 
Bell, N. Eric 
Binford, E. E. 
Booth, V. J. 
Brandon, Joseph F. 
Brewer, Herbert C. 
Briggs, Glen 
Brockson, W. I. 
Bryant, Roy 
Bugby, M. O. 
Clark, Geo. H. 
Clemmer, H. J, 
Cocke, R. P. 
Coe, H. S. 
Conrey, G. W. 
Cramer, W. F. 
Criswell, Judson H. 
Curtis, Harry P. 
Daane, Adrian 
Darst, W. H. 
Davisson. Bert S. 
Deeter, E. B. 



de Werff, H. A. 
Dickson, R. E. 
Dougall, Robert 
Douglas, J. P. 
Downs, E. E. 
Finnell, H. Howard 
Fletcher, C. C. 
Fletcher, O. S. 
Fleming, Frank L. 
Foersterling, H. 
Frank, W. L. 
Freeman, Ray 
Furry, R. L. 
Graham, E. E. 
Gray, Samuel D. 
Gray, W. F. 
Halverson, W. V. 
Hanson, Lewis P. 
Harlan, Harry V. 
Haskell, E. S. 
Hill. C. Edwin 
Hill, Pope R. 
Hodgson, E. R. 
Hodson, Edgar A. 
Hoke, Roy 
Holland, B. B. 



Hotchkiss, W. S. 
Huelskemper, E. H. 
Hulbert, Harold W. 
Hurst, J. B. 
Huston, H. A. 
Jackson, J. W. 
Jackson, L. D. 
Jarrell, J. F. 
Jarvis, Orin W. 
Joslyn, H. L. 
Kelly, E. O. G. 
Kemp, Arnold R. 
Kennedy, P. B. 
Kenworthy, Chester 
Killough, D. T. 
Kime, P. H. 
Kraft, J. H. 
Kuska, J. B. 
*Jacobs, Daniel C. 
Langenbeck, Karl 
Letteer, C. R. 
Lippitt, W. D. 
Longman, O. S. 
*Lund, Viggo 
Mcllvaine, T. C. 
McNess, Geo. T. 



KKi'oK'r oi- i iiK skcui:tarv. 



Mathews, Oscar R. 
Metzger. J. E. 
Miller, Frank R. 
Milton. R. II. 
Morison, A. T. 
Mortimer, Geo. B. 
Movnan, John C. 
Mundell, J. E. 
Murph}^ Henry 
Murray. James 
Nevin, L. B. 
Northrop, Robert S. 
Osenbriig, Albert 
Park. J. B. 



Pate. W. \\ 
Pittman, I). W. 
Purington. James A. 
Ratliffo. Geo. T. 
Reed, Everett P. 
Riley. J. A. 
Robertson, R. B. 
Scluier, Henry W. 
Shinn. E. H. 
Smith, J. B. 
Smith, Howard C. 
Southworth, W, 
Spencer, E. L. 
Taggart, J. G. 

Distribution of Membership. 



Tillman, B. W. 
Torgcrson, E. F, 
Trout, C. E. 
True, Rodney H. 
Van Evera, R. 
Van Nuis. C. S. 
Ward, Wylie R. 
Ware, J. O. 
Wilkins, F. S. 
Willard, C. J. 
Winters, N: E. 
Woo, Moi Lee 



Of the 652 members of the Society, 609 are in the continental United States, 
6 in our island dependencies, 25 in Canada, and 12 in other countries. Of 
the 6 in our dependencies, 2 are in Porto Rico, 3 in Hawaii, and i in Guam. 
The 25 in Canada represent 7 Provinces, 2 in Alberta, 2 in British Columbia, 
5 in Manitoba, i in Nova Scotia, 7 in Ontario, i in Prince Edward Island, 
and 7 in Quebec. Of the 12 in other countries, 2 are in Cuba, i each in Costa 
Rica. Brazil, and Paraguay, 2 in South Africa, 2 in China, 2 in India, and i 
in Japan. 

We have at least 2 members in every state in the Union, and 5 or more in 
31 of the 48 states. The leading state in membership is Illinois, with 36. 
Kansas ranks second, with 33; then follow Texas, 31; Ohio, 29; New York, 
28; California, 22; Iowa, 21; and Indiana, 20. Of the states with less than 
20 and more than 10 members. North Dakota has 19; Georgia, Oklahoma, and 
Wisconsin, 18 each; Minnesota and South Dakota, 15 each; Maryland, 13; 
North Carolina, 12; and Oregon, 11. Colorado, Missouri, and Washington 
have 10 each. In the District of Columbia, where naturally a large part of 
the membership is centered, there are 82 members, excluding and crediting 
to the states a large part of the men who are on field station duty the greater 
part of the year. 

Address List of Members. 

The address list of members this year, if printed, would occupy 13 pages in 
the Journal. For the sake of economy, it seems best to omit this list. 
Changes of address are printed as rapidly as they come to the notice of the 
Secretary, and the names of new members are printed in each issue. These, 
in connection with the address list in the last number of Volume 8, make a 
correct list of the membership. 



5. JOURNAL AND PROCEEDINGS. 

With the beginning of the year, the Journal was changed from a bi-monthly 
to a 9-number basis, the numbers appearing monthly except during June, July, 
and August. While there has been some irregularity in the appearance of the 
monthly issues and some few regrettable errors have been made because of 
unusual and difficult working conditions, on the whole the Secretary and Editor 
believes that the Journal ^las been more satisfactory this year than last. The 
monthly issuance makes more prompt publication possible, particularly of short 



396 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



articles for which quick pubhcity is desired. The total number of pages 
printed during the year is 432, an 8-percent increase over last year and a 
35-percent increase over the volume for 1915. Including those in the December 
number, 40 papers have been printed, with 9 plates and 23 text figures. These 
papers have been contributed by 40 authors, who represent 13 states and the 
District of Columbia. 

The increasing cost of publication, particularly the present high price of 
paper, will make necessary an increase in the income of the Society, if the 
Journal is to be maintained on the present basis or is to make any progress 
whatever. While the Secretary regrets very much to make the announcement, 
the only way to obtain this additional income which seems available is to 
increase the annual dues. It will be necessary also to conduct a vigorous cam- 
paign for new members, for our losses are likely to be far heavier than usual 
and we must have new men to fill the gaps and to increase our total. 

Disposal of Publications. 

In Table i will be found full data on the disposal of Proceedings and Jour- 
nal in the period from October 28, 1916, to October 31, 1917. 

The sale of the publications of previous years has been heavier than usual, 
both the sale of the earlier volumes to those members who desired to complete 
their sets and the sale of complete or partial sets to libraries. The number of 
copies of the current volume distributed to libraries shows a further healthy 
growth, being 88 as compared with 77, an increase of 14 percent. The total 
income from the sale of volumes earlier than the present one was $239.80, as 
compared with $133.10 in the corresponding period last year for one less 
volume. 



Table i. — Data showing the original edition of each volume of the Proceedings 
and Journal, the distribution made previous to and during 1917, and 
the number of copies remaining in stock. 



Edition printed, disposition of 
copies, and number remaining. 


Volumes. 


I 


2 


3 


4 


5 


6 


7 


8 


9 


Edition printed 

Previously accounted for. . 
Distributed to members, 
1917 


501 
364 


517 

373 


516 
402 


514 
411 


750 
638 


750 


750 
575 


750 

668 


1,050 

652 

131 
16 


Distributed to subscribers, 
1917 






























37 


12 
















13 

22 


Sold, 1917 


21 


19 


21 


20 


23 


23 


23 


Total copies distributed . 
Sold on credit orders . . . 


385 

116 
I 


392 

125 
I 


423 

93 
I 


431 

83 
I 


661 

89 
2 


575 

175 

2 


610 
140 


703 

47 
I 


799 

251 
I 


Copies in stock 




124 


92 


82 


87 


173 


140 


46 


250 



y)7 



6. MINUTES OF TllK ANNUAL MEETING. 

Washtngton, D. C, Novkmhkr 12-13, iQi?- 

First Session, Monday Afternoon, November 12. 

The meeting was called to order in the New Ehhitt Hotel at 2 p.m. by Presi- 
dent Jardine and the presentation of papers on the regular program was taken 
up, as follows : 

1. Mineral Food Requirements of the Wheat Plant at Different Stages in Its 
Development (illustrated), by Prof. A. G. McCall. 

2. Effect of Sodium Nitrate Applied at Different Stages on the Yield, Com- 
position, and Quality of Wheat, by Drs. J. Davidson and J. A. Le Clerc (pre- 
sented by Dr. Davidson). 

3. Some Facts Regarding the Soft or Flour Corns (illustrated), by Mr. H. 
Howard Biggar. 

4. Drainage Tanks for Soil Investigations — Some Preliminary Studies, by 
Prof. C. A. Mooers. 

5. Organizing Crop Production on the Basis of the Distribution of the Nat- 
ural Vegetation (illustrated), by Prof. Adolph E. Waller. 

The reading of the papers being concluded, President Jardine announced two 
special committees, as follows: 

Nominating Committee. 
Alfred Atkinson, chairman; F. D. Farrell, and R. W. Thatcher. 

Auditing Committee. 
A. H. Leidigh, chairman, and C. A. Mooers. 

* After announcements by the Secretary regarding the evening session, the 
sessions of the following day, and several other matters of interest, the session 
adjourned. ' 

Second Session, Monday Evening, November 12. 

This session was held jointly with the Society for the Promotion of Agri- 
cultural Science for the presentation of the presidential addresses of the two 
organizations and an address by Dr. Liberty Hyde Bailey, recently returned 
from extended travel in China. It was called to order by Prof. C. E. Thorne, 
presiding, with about 225 persons present. The following papers were pre- 
sented : 

The Outlook in Agricultural Science, by Prof. Herbert Osborn, President 
of the Society for the Promotion of Agricultural Science. 

The Agronomist of the Future, by Dean W. M. Jardine, President of the 
American Society of Agronomy. 

Permanent Agriculture and Democracy (as suggested by the situation in 
China), by Doctor L. H. Bailey. 

After the presentation of these addresses, the session adjourned. 



398 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Third Session, Tuesday Morning, November 13. 

This session was called to order by President Jardine at 9 a.m. In accord- 
ance with the decision of the previous annual meeting, it was devoted to a 
discussion of varietal classification and nomenclature. The following papers 
were presented : 

6. The Classification of Western Wheat Varieties, by Messrs. Carleton R. 
Ball and J. Allen Clark (presented by Mr. Ball). 

7. Naming American Wheat Varieties, by Messrs. Carleton R. Ball and J. 
Allen Clark (presented by Mr. Ball). 

8. Classifying Oat Varieties, by Mr. George Stewart (read by Prof. E. G. 
Montgomery in the absence of the author), 

9. Identification of Varieties of Oats in New York, by Prof. E. G. Mont- 
gomery, 

Professor Montgomery, the chairman of the Society's Committee on Varietal 
Nomenclature, then discussed the general subject briefly, following which Pro- 
fessor Wm. Stuart, Secretary of the Potato Association of America, pledged 
the support of that organization in any movement looking toward the stand- 
ardization of the nomenclature of crop plant varieties. After some further 
discussion, the Society went into business session. 

(Business Session.) 

On motion, the minutes of the last annual meeting as printed in the Journal 
(8: 382-385) for November-December, 1916, were approved. 

The report of the Secretary was read and, on motion, approved. 

The report of the Treasurer was read and, on motion, approved. 

The report of the Auditing Committee was read by the chairman, Mr. 
Leidigh, and was approved, on motion. 

The report of the Executive Committee was read by the Secretary and, on 
motion, approved. 

No report was made by the Committee on Soil Classification and Mapping, 
The report of the Committee on Standardization of Field Experiment's was 

presented by the chairman, Dr. T. L. Lyon, and, on motion, approved. 

No report was made by the Committee on Agronomic Terminology, but a 

brief statement regarding the work of the committee was made by Mr. C, R. 

Ball, a member. 

The report of the Committee on Varietal Nomenclature was read by the 
chairman. Prof, Montgomery, and, on motion, approved. 

The Society then, on motion, adopted the following resolution proposed by 
Chairman Montgomery : 

That the American Society of Agronomy appoint a Committee which shall 
act in cooperation with the American seed trade and any other agencies to 
secure uniformity in rules and practices of varietal nomenclature and regis- 
tration. 

The code of nomenclature proposed by the committee was read and, on 
motion, adopted. 

Brief reports of the local sections in Iowa and Kansas were made by Profs. 
Stevenson and Call, respectively, and reports from the local sections at Cornell 
University, South Dakota State College, and in New England were read by 
the Secretary. 



KKl'OUT 0|- TIIIC Sl-X'KETAKY. 



V)<) 



The report of the Committee' on Nominations was (hen read l)y the Secre- 
tary, as follows : 

President, Dr. T. Lyttleton Lyon, Cornell University. 

First Vice-President, Prof. A. G. McCall, Maryland State College. 

Second Vice-President, Dr. C. B. Lipman, Uniyersity of California. 

Secretary, Mr. P. V. Cardon, U. S. Dept. of Agriculture. 

Treasurer, Prof. George Roherts, Kentucky Agr. Expt. Station. 
It was moved and seconded that the report of the committee he adopted and 
that the Secretary he instructed to cast the unanimous hallot of the Society for 
the nominees. Treasurer Roberts then made a brief statement recalling his 
five years of service and asking that' he be relieved from further duty. He 
pointed out the advantages of having the Secretary and Treasurer located in 
the same city and urged that the two offices be merged. This suggestion was 
approved by the Secretary, who stated that the decision regarding the desira- 
bility of merging the two offices had been arrived at independently by the 
present incumbents. The committee's report was then amended by substituting 
the name of Mr. Cardon for that of Prof. Roberts for Treasurer, the amended 
report adopted, the ballot cast by the Secretary, and the nominess declared 
elected. 

The Secretary then made a brief statement of the financial condition of the 
Society, with the estimated income for 1918 and the probable expenditures 
based on the publication of a volume of the Journal of the same size as that 
printed in 1917. This statement showed that a deficit was almost certain to 
result unless the membership dues were increased. It was then moved and 
seconded that that portion of by-law i which reads " The annual dues for each 
active and associate member shall be $2.00 " be amended by substituting " $2.50 " 
for " $2.00." After a short discussion, the by-law as amended was adopted. 

On motion, the annual dues of those members who enter the military service 
were suspended for the period of that service. 

On motion, the Society expressed its thanks to the management of the New 
Ebbitt Hotel for courtesies extended during its sessions. 

On motion, the Society expressed its appreciation of the services rendered 
by the retiring Secretary and Treasurer during their respective terms of office. 

The Society then adjourned to meet at 2 p.m. 

Fourth Session, Tuesday Afternoon, November 13. 

The session was called to order by Mr. C. R. Ball at 2 p.m., the President 
and both Vice-Presidents being unavoidably absent. The presentation of 
papers was resumed, as follows : 

10. Relation of Weed Growth to Nitric Nitrogen Accumulation in the Soil 
(illustrated), by Profs. L. E. Call and M. C. Sewell (presented by Prof. Call). 

11. Wheat Breeding Ideals, by Prof. Harry Snyder. 

12. Calcium in Its Relation to Plant Nutrition (illustrated), by Dr. R. H. 
True. 

13. Red Rock Wheat and Rosen Rye (illustrated) , by Prof. Frank A. Spragg. 

14. The Triangle System for Fertilizer Experiments (illustrated), by Drs. 
Oswald Schreiner and J. J. Skinner (presented by Dr. Schreiner). 

After an announcement regarding the evening session, the session adjourned. 



400 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Fifth Session, Tuesday Evening, November 13. 

The session was called to order at 8 p.m. by Vice-President Lipman and the 
following papers presented: 

15. Some Tests of an " All-Crops " Soil Inoculum by Mr. Paul Emerson. 

16. Corn and Wheat Soils of the United States, by Prof. C. F. Marbut. 

17. Methods Used in Cereal Investigations at the Cornell Station (illus- 
trated), by Dr. H. H. Love and Mr. W. T. Craig (presented by Dr. Love). 

18. The Significance of the Sulfur in Sulfate of Ammonia Applied to Certain 
Soils, by Dr. C. B. Lipman. 

19. Aluminum as a Factor Influencing the Effect of Acid Soils on Different 
Crops, by Dr. B. L. Hartwell and Mr. F. R. Pember (presented by Dr. Hart- 
well). 

The presentation of papers being concluded, the Society adjourned sine die. 



REPORT OF THE TREASURER. 

From November 13, 1916, to November 12, 1917. 
Receipts. 

Balance on hand, November 13, 1916 $ 458.18 

From C. W. Warburton, Secretary, per his statements Nos. 1-5 780.37 

Membership fees received : 

For 1912, 1914, 1915 (reinstatements) $ 6.00 

For 1916 188.00 

For 1 91 7 964.00 

For 1918 10.00 

Local memberships, 1917 2.00 

Fee for collecting check * .10 1,170.10 

For Journal, Volumes 7 and 8 2.00 

Total receipts $2,410.65 

Disbursements. 

Voucher No. 

E. B. Thompson, rent of stereopticon 107 $ 15 00 

New Era Ptg. Co., printing Journal, 8-5... 108 $192.55 

do. 86-91-2... 121 549.39 

do. 93... 126 174.98 

do. reprints... 127 73.53 

do. 9-^-5... 128 430.90 1,421.35 

L. M. Thayer, printing 109 22.75 

do. Ill 19.35 

do. 117 7-75 

do. 123 2.25 

do. 129 13.00 65.10 

Maurice Joyce Engraving Co., cuts no 12.74 

do. 114 3-68 



I9I6. 




Dec. 


4- 


Dec. 


4- 


1917. 




May 


23. 


July 


16. 


July 


16. 


Aug. 


29. 


1916. 




Dec. 


27. 


1917. 




Jan. 


13- 


Feb. 


16. 


May 


23. 


Oct. 


3- 


Jan. 


2. 


Feb. 


5- 



J^l'TOU'I'S COM M ITTKKS. 



401 



March 


28. 


do. 


. 119 


915 




May 


I. 


do. 


120 


3304 




Oct. 


15- 


do. 


1 30 


12.93 


71-54 


Jan. 


13- 


C. W. Warburton, Si'cri'tar\ 's (.'xpcMiscs . . 


112 


48.90 




March 


12. 


do. 


. 118 


22.20 




Max- 




do. 


122 


21.70 




July 


9- 


do. 


• 125 


17-35 




Nov. 


3- 


do. 


• 131 


43-25 


153-40 


Jan. 


13- 




• 113 




15.00 


Feh. 


5- 




• 115 




4.00 


Feb. 


5- 




116 




6.19 


June 


4. 


Ala. Polytechnic Inst., refund 


. 124 




2.00 












.85 












1,754-43 






Balance November 12, 1917 






656.22 



$2,410.65 

Geo. Roberts, 
Treasurer. 

Auditing Committee's Statement. 

The undersigned have examined the accounts kept by Prof. Geo. Roberts, 
Treasurer of the American Society of Agronomy, and consider them in first- 
class condition. We find the statement of the Society's finances as set forth 
in Prof. Roberts' annual report correct. 

A. H. Leidigh, 

C. A. MOOERS, 

Auditing Committee. 

REPORTS OF COMMITTEES. 

The reports of two special committees, those on audit of accounts and on 
nominations, are found elsewhere. The Auditing Committee's report imme- 
diately follows the report of the Treasurer, while that of the Committee on 
Nominations appears at the proper place in the minutes of the business session. 

The reports of two of the four standing committees of the Society follow 
the report of the Executive Committee. The other two committees, those on 
soil classification and mapping and on agronomic terminology, made no report. 

REPORT OF THE EXECUTIVE COMMITTEE. 

A meeting of the Executive Committee was called at Washington, D. C, 
November 16, 1917, by President Jardine, with all members present. 

On motion, it was decided that the next annual meeting of the Society should 
be held at the same place and on the two days preceding the meeting of the 
American Association of Agricultural Colleges and Experiment Stations. The 
matter of holding a sectional meeting with the Great Plains Cooperative Asso- 
ciation or the Dry Farming Congress was referred to the President and Sec- 
retary, with power to act. 

On motion, the Secretary was instructed to continue as Editor of the Journal 
during the ensuing year. On motion, it was ordered that the Journal be issued 



402 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



monthly during 1917, except during June, July, and August. Decision as to the 
size of the monthly edition and the choice of an editorial board was delegated 
to the Secretary and Editor. 

On motion, the Editor was directed to offer the Journal to senior students 
in agronomy at the various agricultural colleges at a reduced price for one year 
only, this price to be approximately the cost of publication (later fixed at $1.25). 

In March, by a mail vote, the committee unanimously authorized the organi- 
zation of the Illinois Section of the Society, with headquarters at Urbana. In 
May, a local section at the North Carolina A. & M. College was similarly 
authorized. 



REPORT OF THE COMMITTEE ON STANDARDIZATION OF 
FIELD EXPERIMENTS. 

The committee have felt that it is desirable to proceed cautiously in the mat- 
ter of making recommendations regarding methods of field experimentation. 
There are at present very few data on which recommendations could be based, 
and we think that the greatest service that the committee can accomplish at 
this time is to examine carefully the data at hand and to encourage further 
research along this line. We have, therefore, decided to select certain phases 
of the subject, to review the records of experimentation dealing with these, 
and to draw some conclusions if that seems justifiable, but to withhold recom- 
mendations. 

Our efforts have been directed towards the consideration of questions regard- 
ing the use of check plots in field trials and the size of plot that is best adapted 
to obtain reliable results. The former of these has been treated by Professor 
Wiancko and the latter by Dean Jardine and Dr. Lyon, Dean Jardine discussing 
the subject from the standpoint of crop experimentation and Dr. Lyon from 
that of soil experimentation. In addition to this we have compiled a partial 
bibliography of the literature dealing with methods of field experimentation. 
It is hoped that this work on the bibliography will be continued by succeeding 
committees so that the Society will ultimately have a complete bibliography of 
the subject. 

Size of Plots for Field Experiments with Crops. 
w. M. jardine.! 

Taylor (65)2 found that the size of field plots in this country varied from 
one fortieth acre to 2 acres, the smaller ones being most common. The 150 
plots at the Rothamstead, England, experiment station were found to vary in 
size from an eleventh to a half acre in size, the average being a little over a 
fifth acre. He believes " the general proposition may be laid down that the 
size of plots should vary inversely with the degree of uniformity of the soil." 
Except for breeding and selection work the plots should be " large enough to 
allow the unhampered use of ordinary farm machinery upon them." It is 
pointed out that the larger the plots the less is the error of computation. 

1 Dean Jardine desires to state that this review was prepared by Prof. S. C. 
Salmon and to express his appreciation of the service. 

2 Figures in parentheses refer to "Bibliography," p. 415. 



KKI'ORT OF rOMMITTKE ON KIKI.D K\ I'l'.RI M ICNTS. 



" For most tests one tenth of an acre should he considered ahout tlie mini- 
mum size while the maximum would he limited largely hy the extent of land 
available." "One-tenth-acre plots are most convenient for conii)ulinj^^ results 
and the area is ample for the use of most machinery." 

He concludes that "the size of plots is largely a matter of convenience and 
local conditions." The best size " depends first upon the area of land available, 
second upon the uniformity of the soil, and third upon the num])er of plots 
necessary for a given experiment." 

Smith (59) points out that in tests with corn the plots should be large enough 
to permit a fair representation of the seed under investigation. This is stated 
to vary with the nature of the test and no definite rule that will apply to all 
cases can be laid down. At the Illinois station the single or 2-row plots for 
variety tests have been abandoned because they seem to be too narrow for a 
fair test. Tall varieties shade short varieties and plants with extra strong 
foraging powers have an advantage over a weaker adjoining variety. The 
system has been adopted of planting the variety plots 5 rows wide and discard- 
ing at harvest time the 2 outside rows of every plot. 

Cory (13) compared row plantings and field plots for variety tests in experi- 
ment's at Dallas, Texas, and McPherson, Kans. He noted the following advan- 
tages for the row plantings: (a) Easier to get help for harvesting the row 
plantings than to secure teams to harvest the field plots; (b) the bundles from 
rows can be stored and thus prevent damage from bleaching, sprouting, shat- 
tering, etc. ; (c) all short and weak stems are gathered better than in field plots ; 
and (d) economy in land and labor. He gives it as his opinion that except for 
" varieties with short or weak stems, or in cases of unequal germination, insect 
infestations, or errors made in threshing, . . . the row plantings will furnish 
an accurate check for field plots." 

In tests conducted at the Cornell Agricultural Experiment Station, Lyon (39) 
found that the mean deviation from the normal yield is considerably less for 
plantings in 17-ft. rows when 10 such rows represent a single test than for 
tenth-acre plots when each plot represents a different test. " The advantage 
from the small plots is not only in point of accuracy but also in the area of 
land required. Seven one-tenth-acre plots covered an area of 30,492 square 
feet while seventy of the rows required only 1,190 square feet. The use of the 
row method in variety testing is commended by the results of this test." 

Piper and Stevenson (55) recom^mend the use of fortieth-acre to tenth-acre 
plot's as the minimum for varietal and similar tests with small grain, these 
plots to be replicated at least twice and preferably five times. For row tests, 
rows I rod or more in length and 6 to 10 inches apart are recommended. For 
varietal and similar tests with corn, the minimum standard recommended is 5 
rows each of 25 hills or 5 rods long, the outer two rows to be discarded. 

Lyon (40), from a review of experiments dealing with experimental error in 
field plots and from his own experiments and those of his associates at the Cor- 
nell station, concluded that " There seems to be little gained in accuracy by 
using plots larger than one fiftieth of an acre in size, when the yield of crops 
i>: made the criterion. . . . For accuracy in sarnpling soil, the smaller the plot 
the better. When it is possible to give small plots careful treatment and to 
replicate each treatment several times, the smaller plots are preferred to large 
ones. ... An area of one twenty-fifth acre of land in four widely separated 
plots devoted to any one test secures a much larger degree of accuracy than the 
same area of land in one body." 



404 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Mercer and Hall (43) grew an acre of mangolds which they harvested in 
200 sections of one two-hundredths acre each and i acre of wheat which they 
harvested in 500 sections of one five-hundredth acre each. By combining these 
plots into larger plots and comparing the probable errors of the different sizes 
so obtained they were able to arrive at a comparison of the accuracy of tests 
conducted with different-sized plots. They concluded that the error dimxinishes 
with the size of the plot but the reduction is small when the plot grows above 
one fortieth acre. For practical purposes the authors recommended that in 
any field experiment each unit of comparison (variety, method of manuring, 
etc., according to the subject of the experiment) should be given five plots of 
one fortieth of an acre each, systematically distributed within the experimental 
area. 

Wood and Stratton (78) investigated the error of field experiments by two 
independent methods. By the first method an apparently uniform area of 
about an acre was marked out in the middle of a field of mangels arid divided 
into plots of one thousandth acre each. These were harvested and weighed 
separately. A total of 1,050 plots were harvested, but 150 on one side of the 
area were excluded because of markedly lower yields. These small areas were 
combined into larger ones and the probable error for different combinations 
determined. By the second method, duplicate pairs of plots in experiments 
reported in various publicati'ons were used. For this purpose the probable 
error of 400 pairs of plots, including experiments with wheat, barley, oats, 
mangels, swedes, and potatoes and varying in size from an eightieth to a half 
acre was determined. The probable error varied from 3.1 to 7 percent of the 
mean except for the thousandth-acre plots of mangels, for which it was 12 
percent. The authors conclude that the probable error is "-independent of the 
size of the plot employed provided this is one eightieth of an acre or larger." 

Olmstead (52) reviewed the experiments of Mercer and Hall, Montgomery, 
and Lyon, applying to their data the method of least squares. He concluded 
that when wheat is grown, a number of small plots, ranging down to 0.0007 
acre, is much better than the same total area in a single plot. In his opinion 
the data indicate that one large plot is better than one small one but no opti- 
mum size is indicated. He concludes that " the estimation of the probable 
errors of a large number of small duplicate plots well distributed in the area 
devoted to a field experiment indicates that the precision of agricultural experi- 
ments can be increased by replicating the experiments on small plots. 

" Replication is urged not only for giving greater precision in experiments in 
which the results justify the increased labor but also for furnishing a means 
whereby the experimenter and his reading public can determine what reliance 
can be placed upon the data and conclusions of the experiment. 

" Replication also lessens the uncertainties involved in the use of check plots 
and may decrease the total area required for the conduct of field experiments." 

From observations in variety and other field tests conducted at the experi- 
ment farm at Newell, S. D., Salmon (57) concluded that greater accuracy in 
variety tests can be secured by dividing the tenth-acre plots (33 X 132 ft. plus 
5-ft. alleys) into five plots of approximately one fiftieth acre each (6 X 132 ft.) 
separated by alleys of 1.6 ft., and replicating each variety five times. By this 
method the same area of ground was required for a given number of varieties 
and much greater accuracy was secured with an additional expense for labor 
in seeding, harvesting, and thrashing of less than 50 percent. 



Ri-.roKT ()!• COM M ri ri-:!-: on iM;in m icnts. 



T>:irl)or (4) pi)intccl out that there are proportionately more phmts alon^ t'l^ 
border in small plots than in lari>e ones and since the plants on the outside of 
the ])U)ts are more productix'e than those within the i)lot, the error from this 
source is greater in small plots than in larj^c ones where tlie plots are separated 
by cultivated or uncropped pathvva>'S. lie {^ives no experimental data U) show 
the optimum si/.e of plots. 

Mortenson (51) believes tjiat from 0.005 to 0.001 hectare (about 535 s(|. ft.) 
is the proper size of plots and states that better results have been obtained from 
small plots repeated often than from larger plots. 

Montgomery (48) grew a plot of wheat 77 X 88 feet and harvested it in 224 
separate plots each 5.5 feet square. The soil was of average uniformity and 
fertility, producing on the average about 25 bushels per acre. These small plots 
w^ere combined into larger plots and the error for the different sizes so obtained 
was computed. Jt was found that "increasing the size of plot above four 
adjacent blocks has had very little effect on reducing variation." He states 
that plots 5.5 X 16 ft. is a very convenient size. With plots of this size "48 
comparative tests per acre can be made, repeating each ten times, with a much 
higher degree of accuracy than when tenth-acre plot's are repeated twice or 
three times." 

In another paper (49) he concludes that "to increase the size of the block 
up to a certain limit rapidly decreases variability; but error cannot be indefi- 
nitely decreased by continuing to increase the size of the plot. ... To increase 
the length of the row four times decreases the deviations about one half. By 
increasing the length of the row or the size of the block the number of repeti- 
tions necessary is decreased, but the total area required to secure the same accu- 
racy is increased. An excellent size where land is plenty would be 2 to 4 rods 
in length for rows and 5 X 16 feet in area for blocks." 

Size of Plots for Field Experiments with Soils. 

T. L. LYON. 

Field experiments with crops are concerned with the soil merely as a means 
of producing the crops. Uniformity in the soil is the great desideratum, 
although other factors, such as the degree of fertility or acidity, may affect the 
results. As these latter factors may best be gauged by repeating the experi- 
ments on other soils, the problem in the end resolves itself into one of secur- 
ing a uniform medium for any set of trials. The results are measured only in 
terms of crop yields or quality. 

Field experiments with soils are as dependent for their success on the uni- 
formity of the soil at the beginning of the experiment as are experiments with 
crops, but after the experiment is once begun the uniformity is necessarily 
destroyed. The results of the experiment may be measured in terms of crops 
or in terms of the soil itself, which is generally accomplished by the use of 
chemical, physical, or biological tests. 

As both crop and soil experiments demand a uniform soil at the outset, any 
discussion of the subject from either standpoint would naturally cover much 
the same ground. Crop experimentation, however, includes a discussion of the 
use of single rows of plants, which is obviously unsuited to soil experiments. 
Experimentation with soils calls for the discussion of measurements of the 
uniformity of soil plots by other than cropping tests. Unfortunately this latter 



4o6 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



phase of the subject seems not to have received much consideration at the hands 
of experimenters. 

The function of this paper is to review the pubUshed investigations that are 
available on the subject of the relation of the size of field plot to the experi- 
mental error introduced by lack of uniformity of the soil. The review is con- 
fined to reports of experiments and does not pretend to be exhaustive. 

Any discussion of the degree of accuracy to be secured by the use of plots 
of different sizes will naturally carry with it a consideration of the effect of 
replication of treatments on scattered plots," as that gives an opportunity to 
consider the relative effectiveness of a given area of land in one body or in 
scattered units. 

Wood and Stratton (78) have calculated the probable errors involved in the 
use of plots of various sizes by taking the yields of 400 duplicate plots used in 
many different experiments and planted to several kinds of crops. For each 
pair of similar plots they have calculated the mean, and the difference between 
each ]j!ot and the mean. Using this as a basis they have calculated the probable 
error tor plots of one half, one quarter, one twentieth, one fortieth, and one 



eightieth acres, using the formula p.e. = o.Gy^j , in which 6^ equals the 

^ n(w — i) 

sum of the squares of the deviations from the mean and n equals the number of 
tests. The probable error for each size of plot is as follows : 

45 pairs of plots each acre 3.5 percent. 

52 pairs of plots each i/4 acre 3.5 percent. 

29 pairs of plots each Yoq acre 3.9 percent. 

200 pairs of plots each y^o acre 4.6 percent. 

75 pairs of plots each %o acre 3.1 percent. 

There appears, when judged by these results, to be no advantage in using 
plots of more than one eightieth acre. 



Wood and Stratton also used another method for estimating the probable 
error arising from the use of plots of different sizes. In an apparently uni- 
form field of beets an area of about i acre was measured off and divided into 
plots of 0.001 acre each, and at harvest each plot was weighed. In all 1.050 
plots were weighed, but it was decided to reject 150 on account of their apparent 
lack of uniformity. The values of the probable error of these plots calculated 
by the method used previously is as follows : 



Number of plots. Area. Pr bable error. 

900 %ooo acre 12 percent. 

36 Mo acre 4 percent. 

25 1/^8 acre 7 percent. 

25 %o acre 5 percent. 



If we consider the results obtained by both methods and discard the results 
from the thousandth-acre plots, which are obviously too small for soil experi- 
ments in the field, we have left eight independent results for the probable error 
in field experiments, which range from 3 to 7 percent. It may be deduced 
from this that the probable error of field experiments carefully conducted is 
about 5 percent. It is also interesting to note that between plot's of a half acre 
and those of an eightieth acre the probable error is independent of the area 
of the plot. 



KKI'ORT OK COMMITTKK ON KIKI.I) KX I'1:R I M ICNTS. 



407 



The authors then show by the tlieory of probal)iHties that with a prohahlc 
error of 5 pereent the least significant difference between two plots is 19 percent. 
In other words, differences of less than 19 i)ercent in yield would not justify 
conclusions as to relative effects of different treatment. In order to reduce 
th^ significant differences they advise replicating the plots and proceed to 
develop mathematically a table showing the number of replications necessary 
to obtain a certain precision in percentage difference between yields. Thus for 
a significant difference of 10 percent, four plots arc necessary; for a difference 
of 6 percent, ten plots are necessary. 

Mercer and Hall (43) also determined the deviations which appear in the 
use of plots of different areas. The areas used were selected from fields which 
seemed to the eye to be uniform. One of these fields was in wheat, and the 
other in beets. The authors employ the method of least squares for calculating 
the errors from plots of different sizes. The plots vary in size from one two 
hundred and fiftieth to one twentieth acre, from the study of which they con- 
clude that the error diminishes as the size of the plot increases, but not propor- 
tionately, and that there is little advantage in using plots of more than one 
fortieth acre in area. They also conclude that it is desirable to replicate the 
plot treatments, but that there is not much to be gained by increasing the 
number of replica plots beyond five. In both of these conclusions they are 
somewhat at variance with Wood and Stratton. 

Hall and Russell (26) divided a wheat field at the Rotbamstead Experiment 
Station into plots of one five-hundredth acre each and harvested and thrashed 
each plot separately. The yields of grain averaged about 4 pounds per plot. 
These, when grouped at intervals of 0.2 pound produced a frequency curve 
that indicated a fair degree of uniformity. When the probable error for any 
one plot was calculated for single plots and for groups of plots, making plots 
of various sizes up to one tenth acre, the results were as follows : 

Size of plot. Probable error. 

1/^00 acre : 7.8 percent. 

•/^50 acre 6.7 percent. 

yi25 acre 6.0 percent. 

y^o acre 4.2 percent. 

acre 3.8 percent. 

i/4o acre 3.4 percent. 

The authors conclude from these figures that there is but litt^i - 'vantage, in 
point of accuracy, to be gained by the use of plots larger than one riftieth acre 
in size. They then calculated the probable error for units of five plots scat- 
tered over the area under experiment, which reduced the error to 2.4 percent, 
from which they conclude that each treatment should be repeated five times 
w4th plots one fiftieth acre in size in order to reach the greatest accuracy com- 
patible with due economy of land and labor. 

Montgomery (48) divided an area of wheat into 224 plot's, making each plot 
5.5 feet square. Each plot was harvested and thrashed separately. By com- 
bining the yields of several contiguous plots larger plots may be formed and 
the accuracy of larger and smaller plots may thus be compared. In this way 
plots of y-m acre, %o acre, acre, and % acre were formed and the coefficient 
of variability calculated from yields of wheat for three years with the follow- 
ing result : 



4o8 



JOURNAL OF THE AMERICAN SOCIETY 



AGRO; MY. 



Size of plot. ^^o^fiScie of variability. 

3/144 acre 14.66 1 ercent. 

YsQ acre ' 8.95' percent. 

Yig acre . . . 8.16 percent. 

Yq acre 7.85 percent. 

These figures indicate but little gain in accuracy by increasing the size of 
the plot above Vsq acre. On the other hand if instead of increasing the number 
of contiguous plots a corresponding area be secured by scattering the plots 
systematically over the area used in the experiment the accuracy is greatly 
increased, as shown by the following statement: 

Number of plots. Area thus secured. Coefficient of variability. 

I 1/444 acre 14.66 percent. 

4 YsQ acre 5.51 percent. 

8 Yis acre 3.22 percent. 

16 Yd acre 2.01 percent. 

In experiments by Lyon (40), 37 hundredth-acre plots were used. These 
were planted to maize and all received the same treatment. The yields when 
plotted in a frequency curve indicated that the figures were suitable for calcu- 
lating the probable error which may be expected to occur on any plot in the 
series. Calculated in the customary way the probable error for any hundredth- 
acre plot was 5.2 percent. In order to compare the hundredth-acre plots with 
larger ones the plots were grouped in four contiguous plots, thus giving an 
area of 0.04 acre. Calculated in a similar manner the probable error for any 
0.04 acre plot was 4.5 percent. When the o.oi acre plots were arranged so 
that every fourth plot was a check and in units of four plot's scattered over the 
area under experiment the probable error for any treatment was reduced to 2.0 
percent. 

In the same paper a report is given of an experiment in which a field of pota- 
toes was harvested by rows, each row constituting a plot of 0.02 acre. By com- 
bining these, larger plots were formed. When the data of the yields is handled 
in the same way as the preceding the probable errors for plots of different 
sizes are as follows : 

Size of plot. Probable error. ^ 

Y50 acre 6.1 percent. 

Y25 acre 4.8 percent. 

%5 acre 3.8 percent. 

%5 acre 3.2 percent. 

%5 acre 3 3 percent. 

In this field there is no significant increase in the probable error until the 
plot's become as small as one twenty-fifth of an acre. 

Holtsmark and Larsen (32) proceeded in a somewhat different way to esti-^ 
mate the errors arising from the use of plots of different sizes. They first 
determine what they call the true error, which is found by planting the entire^ 
area, consisting of all the plots to be tested, with a certain crop. The average 
for the yields of all of the plots is termed the " true value " of the plots, and 
the difference between this average and the actual yield of any plot is termed the 



R>"'HbRT (^i-"'' ' I ri'i:i-: on i:\i'i:i<i mknts. 



"true error." If'lic " iriR' errors" arc used as the basis for calculating the 
"mean error "*by means of the following formula: 

II ' 

in which / 'i, / ';... • •• / '„ arc the true errors, and )i is the number of errors of 
which the mean error is to be determined. They proceed on the theory of 
probabilities that when n quantities vary from a certain value through faulty 
observation, and when the average error of the separate quantities is m, the 

m 

error of the arithmetic mean of the quantities will 1)C + — =. 

V n 

If instead of one plot for each fertilizer or other test, two plots are used, 

the mean error becomes = 0.71 ; if three plots are used the mean error is 
I 2 

—^=0.58; four plots, — m 0.50, as great as the mean error when only one 
I 3 V4 
plot is used. 

By experiment the authors found these values actually to be i : 0.658 : 0.418 
instead of i : 0.71 : 0.50. They ascribe the discrepancies to the fact that the 
errors are not all accidental, as is assumed by the theory of probabilities, but 
that there w^ere also errors caused by non-uniformity of the field. 

The above explanation is necessary in order that the reader shall understand 
their method of estimating the comparative accuracy of plots of different sizes. 
A field with a crop growing on it was divided into 96 plots of Yiq acre, and 
these grouped to make 48 plots of % acre, 24 plots of ^ acre, 12 plots of % 
acre, and 6 plots of i acre. The mean error for each group of plots was then 
calculated with the following results : 



Mean error in percent- 
Number of Size of Mean age of average 
plots. plots. error. yields. 

96 V16 acre 0.887 i7-4 

48 % acre 1.610 15-8 

24 % acre 2.910 14.6 

12 y2 acre 5.180 12.7 

6 I acre 9-390 11.5 



As compared with the other experiments cited there is a relatively larger 
error with the use of small than with large plots. These results are of interest 
chiefly because of the method used in securing them. 

Olmstead (52) has worked over the data recorded by Mercer and Hall, Mont- 
gomery, and Lyon, and has calculated the probable error for any plot in Mercer 
and Hall's wheat and mangold fields, Montgomery's 5.5 ft. square wheat plots, 
.md Lyon's potato rows, when the unit areas in these fields are grouped con- 
iguously and when scattered over the field used in the experiment. He shows 
that all of these data indicate that the precision of field experiments may be 
increased by replicating the tests on small plots down to 0.0007 acre rather 
than by using the same area in one large plot. He says : " When many small, 
well-distributed duplicate plots are used, soil of all degrees of productivity is 
likely to be found in each group of plots, and this is the main argument for 
replication in field and pot experiments." 



410 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



An attempt has been made by Kaserer (35) to ascertain the number of bor- 
ings that are required to adequately estimate the composition of the soil of 
a field plot. Two plots were selected, each of which had a known history ex- 
tending over a number of years previous to making the test. The previous 
treatments had been uniform over the entire area of the plots. Each plot was 
164 feet long and 65.5 feet wide. Samples were taken to a depth of 10 inches 
by means of a Kopetzky borer. The borings were made on the line of the 
diagonals of the plots and there were nine borings in all. As the borings were 
equidistant on the diagonals they would be 29.5 feet apart. 

The sample obtained by each boring was analyzed separately, determinations 
of dry substance, total nitrogen, and nitrate nitrogen being made in duplicate. 
While the duplicate determinations of the same sample agreed excellently well 
there were larger differences between the samples. This leads him to call atten- 
tion to the great difficulty in getting a representative sample of the soil of a field 
plot and to recommend that borings should be taken on at least every square 
meter of surface. 

An attempt was made by Allison and Coleman (i) to measure the biological 
uniformity of two plots of land. For this purpose 2 twentieth-acre plots were 
selected, one of which was in timothy and the other had produced a crop of 
corn previous to the test. The former was a heavy clay and the latter a sandy 
loam. Ammonification tests with different nitrogen carriers were made with 
soil samples taken at several points in each plot. The authors conclude that 
" where plot's are uniform in character the biological variations of the soil at 
different points in the plot are not great, or else we are not able to detect these 
differences by the present methods." 

The Use of Check Plots in Field Experiments, 
a. t. wiancko. 

Although there is a considerable amount of literature on various phases of 
field plot experimental work with crops and soils, very little has been published 
concerning the -use of check plots. Many of the articles that do touch upon 
the subject deal primarily with the application of calculation methods in reduc- 
ing the experimental error in the yields secured and only incidentally refer to 
check plot's as such. The list of references thus far gathered by the Com- 
mittee is doubtless incomplete and a full review of the literature dealing in 
any way with check plots or checking systerhs is not possible at this time. 
Some of the references known to the Committee could not be secured for 
examination and review. In the paragraphs below an attempt has been made 
briefly to summarize the principal points in the articles referring to check plots 
and checking systems involving check plots of one kind or another. 

Atwater (2) calls attention to the unevenness of the soil and the fact that 
duplicate trials seldom agree, and suggests the preliminary testing of areas 
intended for experimental purposes by uniform treatment for a series of years 
and using only such as prove to be intrinsically uniform. Where experiments 
are to be undertaken at once and preliminary testing is not feasible, he suggests 
the use of small plots several times repeated. In another part of the discussion 
he suggests duplicating manured plots and using not more than two* or three 
unmanured plots for ordinary experiments. 

In a later report (3), which gives full instructions for conducting field soil 



RKPOUT ()!• COM M 1'I"1'I:K ON VWAA) ICX i'l':RI M KNTS. 



411 



fertility tests. Atwater says with reference to check plots: " It is of the greatest 
importance that several uninanured plots he left for comparison. For eight 
manured plots, two unmanured will siiflice ; hut where there arc more than that, 
three, one at each end and one in the middle, or, if the numhcr is large, one in 
the middle and one half-way between this and each side would be advisable. 
Vou will have very little idea how uneven an a])parently uniform soil may be 
until you make the trial." 

Morgan (50) reports a study of the variation in yields on different areas in 
similarly treated fields of corn and wheat which were divided into a large 
number of small plots. By calculations based on assumed checks at a number 
of dil?erent intervals, it is shown that the average variation is constantly re- 
duced with the frequency of the checks. 

In his book on fertilizers and manures. Hall (24) briefly discusses the layout 
and management of experiment fields and recommends repeated smaller plot's 
as being preferable to larger single ones. He further calls attention to a plan 
used by Dr. Sonne in Denmark, involving four repetitions of each treatment 
distributed around the field, as a means of reducing error. 

Thorne (66), in a circular which contains detailed instructions for conduct- 
ing field experiments, states regarding check plots : " A matter of great impor- 
tance, too often lost sight of in field experiments, is the repetition of check 
plots. In the most uniform soils there will be some variation in the produce 
of adjoining plots from season to season. Even were the actual plant food the 
same, the variations in level which occur on all soils will produce an unequal 
distribution of moisture, and moisture may often be a more important factor 
in determining crop yield than plant food. The ideal system of plot experiment 
would leave every alternate plot as a check. Next to this comes the plan of 
leaving every third plot as a check, thus having a check plot on one side or the 
other of every plot under treatment. In fertilizer tests the check plots may be 
unfertilized or subjected to uniform dressings with a standard fertilizer or 
manure, depending upon the object of the experiment. In variety tests the 
check plots should be planted to a standard variety." In discussing the calcu- 
lation of the increase, the author recommends the method of comparing the 
treated plot yield with the normal yield calculated from the two nearest checks, 
assuming the difference between the checks to be uniformly progressive. 

Gardner and Runk (21), in connection with a report on a variety test of oats 
at the Pennsylvania station in 1908, discuss the checking method using a stand- 
ard variety every third plot and computing the yield of the intervening varieties 
from a calculated check yield, assuming a uniform change in the soil from one 
check plot to the next. 

L3-on (39) reports the comparison of wheat variety tests conducted by means 
of tenth-acre plots with every third plot a check, and 17-foot rows with every 
tenth row a check. The probable error for each checking system is calculated 
and found to be smaller in the row system. It is concluded that the row method 
may be preferable in point of accuracy as w^ell as in requiring less space. 

Lyon (40), in a report on some experiments to estimate the error in field 
plot tests, discusses the use of check plots at some length, explaining their use 
as being based upon the assumption " that the productiveness of the soil changes 
gradually from point to point and that consequently the natural productive 
capacity of any plat may be determined by its distance from two plats on 
either side, the natural productive capacities of which are known." Experi- 



412 



JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



ments are cited, showing that this is not always true. It is brought out, how- 
ever, that the more frequently check plots are introduced the more accurate 
will be the results. Different methods of making corrections by means of 
check plots and the use of repeated series as means of reducing error are dis- 
cussed. The conclusion is reached, regarding the distribution of checks, that 
" the use of check plats every second or third plot secures greater accuracy than 
when no checks are used," but the data at hand do not indicate any advantage 
from the use of check plats at less frequent intervals. 

Piper and Stevenson (55) discuss the various factors affecting the problem 
of field experiments in crop and soil studies and suggest minimum standards 
for each of several lines of investigation. As regards checking, different sys- 
tems are recommended for different lines of work. For corn variety tests, 
every fifth plot or fifth row as checks and five replications in rows and two 
to three replications in plots are recommended. For small grain tests, every 
fifth row or third plot as a check and from two to five replications in plots 
and ten in rows are recommended. For soil fertility tests, it is recommended 
that every third plot be used as a check and that the whole series of plots be 
repeated as many times as may be required for the growing of each crop in the 
test every year. 

In an experiment to determine the experimental error in field tests, Mercer 
and Hall (43) found wide variations in the yields in different parts of an 
apparently uniform field and conclude that a single season's results on single 
plots may be very unreliable and that trials should be replicated. Five repli- 
cations are recommended to bring the error within 2 percent. Mention is also 
made of the fact that variations due to season can be checked only by repeating 
over a series of years. 

Mitscherlich (46), in a study of the experimental error in twenty experi- 
ments, found that the error is lessened by reducing the size and increasing the 
number of plots (duplicate as well as control). 

Montgomery (48) found that repeating ten times in small plots was much 
more accurate than large plots repeated two or three times and concludes that 
the only method of securing comparative yield tests that will meet all of the 
fluctuating variations is systematic repetition. 

Montgomery (49) presents a considerable amount of experimental data on 
different systems of planting and checking in field studies with wheat. He 
concludes that systematic repetitions is the best way to reduce error and that 
ten to twenty repetitions should be made to insure a satisfactory degree of 
accuracy. It is also known that while having every other row a check gives 
the highest degree of accuracy, the total number of plots required for the 
same degree of accuracy is greater by this method than by systematic repetitions. 

In a study of the size of plots and effect of repetition on accuracy, Mortensen 
(51) shows that better results have been secured from small plots repeated 
often than from larger single plots. It is concluded that the number of repe- 
titions should be from eight to ten ordinarily, depending upon the number of 
factors to be determined by the experiment. If more than two factors are 
wanted, more than ten plots should be used. 

Hanroth (26) discusses methods of reckoning the variation of each plat of 
the series from the mean of their yields. The Gauss formula is applied to 
various fertilizer experiments as a means of making corrections for the observed 
variations. 



Ri-.pouT oi- COM M iT'n:i': ox i:x i'i;ui m icnts. 



413 



Salmon (57) (liscussi.>s i1k' iuhmI of greater accuracy in field plot experiments 
and with particnlar rcferenco to variety tests suKRests the use of the small plot 
repeated several times. 

Salmon (58) presents some data from a variety test of harley in 1912 which 
was non-uniformly affected by some unknown cause, probably uneven drifting 
of snow, and shows that no fixed check plot system of correction could be 
safely applied to the results. 

Olmstead (52) presents a number of different sets of data to illustrate the 
application of the method of least squares in computing the results of field 
plot experiments to a comparable basis and discusses the use of check plots 
and repeated plantings. In the discussion of the check plots, mention is made 
of the practice of having one third to one fifth of the plots as checks and it 
is suggested that a combination of the methods of comparing with a normal 
yield based on the two nearest checks and the method of comparing with the 
average yield of all checks be used in calculating the results to a comparable 
basis. From the experimental areas of the study, including several different 
crops, it is concluded that replication of experiments on small plots is tlie best 
means of increasing accuracy. 

Wiancko (76), after a brief discussion of the use and distribution of check 
plots in existing soil fertility experiment fields and calling attention to the non- 
uniformity of the annual curve of the check plot yields, points out the fact 
that the check plot as usually employed, having all the produce removed from 
it and nothing returned, becomes poorer and poorer and that during the process 
of reduction of fertility the initial relations of one check plot to another may 
change and that finally the plots become so poor that crops either do not 
develop normall}^ at all or are subject to uneven annual fluctuation caused by 
various unfavorable climatic conditions, plant diseases, and insect injuries to 
which their unthriftiness makes these plots especially subject. The question 
is then raised as to whether it would not be wise to maintain the check plots 
in a reasonable state of fertility by means of a uniform manurial treatment 
calculated to produce at least fair yields of all the crops in the rotation. 

Pritchard (56) presents data from sugar-beet experiments showing the effects 
of check rows at different distances and of repeated plantings. It is shown 
that both frequent checks and several replications should be used to secure 
reasonable accuracy. 

Stockberger (63) applies different methods of calculation to the reduction 
of error in connection with a test of thirt}' rows of hops and concludes that 
the methods are of little value, as fresh errors may be produced. These 
errors may be reduced by corrections for imperfect stand and by replication. 

Surface and Pearl (64) discuss the variability in the fertility of the soil as a 
factor in field experiments and propose a method of calculation to correct for 
such errors with special reference to crop variety tests. The method involves 
the calculation by the contingency method of the probable yield of each plot 
on the supposition that they have all been planted with a hypothetical variety 
whose mean yield is the same as the observed means of the field. This calcu- 
lated yield is then used to correct the actual yield of the plot to the mean of 
the field. The ordinary check plot method is not considered satisfactory. 



414 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 



Conclusions. 

The investigations here reviewed seem to admit of the following conclusions: 

The probable error for any one plot decreases as the size of the plot increases. 

For any given area devoted to a test of a single treatment the probable error 
is less if this area be divided into sub-areas and scattered over a field than if 
the treatment be applied to one single body of soil. Furthermore the probable 
error appears to decrease with increased subdivision of the area down to very 
small units (possibly 0.001 acre). 

For the same reason greater accuracy may be secured by using scattered 
treatments on small plots than by one treatment on a large plot even when the 
small plots do not cover as much land as does the large plot, but the ratios 
involved in this conclusion have not been definitely worked out. 

Experimental evidence appears to leave no doubt as to the possible accuracy 
of replicated row plantings as compared with single field plots when equal 
areas are used for each, and even when smaller areas are used for the row 
plantings, but here again no definite ratios are available. 

While there are very few data on the frequency of borings necessary to 
secure accuracy in sampling field plots, such data as are available indicate that 
it is difficult to secure an accurate sample and that borings should be taken at 
close intervals of space. 

Experiments in the use of check plots seem to indicate that the usefulness of 
these plots increases with the frequency with which they are distributed among 
the test plots. There are not sufficient data, however, to guide one in estimat- 
ing how frequent their use should be in order to obtain any desired degree of 
accuracy. 

Suggestions. 

The committee do not desire to recommend to the Society that it adopt any 
particular size of plot, number of replications, or method of handling check 
plots. We feel that the subject has not been sufficiently developed to make it 
advisable for the Society to place itself on record with respect to any one 
procedure as opposed to the many possible ones. We strongly urge, however, 
that investigation of the whole subject of field plot trials, and particularly that 
phase of it that concerns the size of plots and handling of check plots, be con- 
ducted- by members of the Society. 

More data should be secured on the probable error involved in the use of 
plot's of different sizes and the number of replications required to reduce the 
error to any desired degree of accuracy. With a sufficient amount of such 
data at hand it will be possible for the experimenter to calculate exactly how 
many replications he must have with plots of a given size in order to get sig- 
nificant differences in yield of crops. It is quite evident that if an experiment 
is not conducted with sufficient accuracy it is worthless and it would have been 
better not to have attempted it. If, as stated by Wood and Stratton, an experi- 
ment on single tenth-acre plots requires differences in yields of 20 percent in 
order to insure that the differences are due to the treatments, it is very evident 
that a considerable number of our field plot tests do not admit of the interpre- 
tations that have been placed upon them. 

The relative accuracy of row plantings and plots is a subject that requires 
investigation. The practice of row plantings seems to be subject to error 
from the effect of adjacent rows that contain more vigorous or less vigorous 
plants. How to meet this difficulty is a problem for the experimenter. 



RKI'OUT Ol' COMMITTKK ON I'll-.I,!) KX I'l'.U I M NTS. 



Very little work lias been done on securing representative samples of soil 
from field plots. Refinements of analytieal methods arc nseless if the samples 
do not represent tlu' ;i\craj;c' soil of ihe plot within at least rather narrow 
limits of error. 

In the study of the use of check plots much experimentation must be done 
before we shall know how best to scatter and treat these plots. There is the 
question whether checks shall be used only in the same number and manner as 
the replicated test plots, or whether they shall be used more frequently. If 
they are to be used more frequently, at what intervals shall they be distributed? 
What method shall be used for calculating the yields of test plots from the 
checks? Finally, there is the important question regarding the maintenance of 
fertility of the check plots in order to make a fair comparison with the test plots. 

All of these problems must be studied. They cannot be decided by vote of 
the Societ3^ But this Society represents the men who are to solve these prob- 
lems if this country is to do it. Let us then devote our energies to this work 
and let every experiment station that is conducting field experiments use some 
of its land and resources to investigate at least one phase of the subject during 
the coming year. 

It is not a simple matter for a committee to arrange for a systematic plan of 
cooperative experimentation with methods of field experimentation. The 
equipment required is too elaborate. There is probably no time when many 
experiment stations could conduct experiments on any one phase of the sub- 
ject, except possibly on the number of borings required to adequately represent 
the soil of a plot. The matter cannot be handled like a chemical analysis for 
which almost any laboratory is equipped. 

It seems likely that it will be necessary to leave the work to voluntary effort 
on the part of agronomists, but the Society can continue to do much to stimu- 
late activity in this direction. The present interest should not be allowed to 
subside. It is through the Journal of the Society, discussions at the meetings, 
and work of the Committee on Standardization of Methods that this Society 
can operate. All of these agencies should be used to the fullest extent. The 
committee feels that the agronomists of the country are alive to the necessity 
of improving methods of field experimentation and that by continuing the 
active participation of the Society in the work a mass of data will be collected 
that will eventually make it possible to compute the experimental error for any 
desired procedure and thus to insure the accuracy of the results. 

Bibliography. 

The Committee realizes that the bibliography which follows is not complete, 
but publishes it here as a report of progress. Those interested in the stand- 
ardization of field experiments are urged to send to the Committee any addi- 
tional titles which may come to their notice. 

1. Allison, F. E., and Coleman, D. A. Bi(51ogical variations in soil plots as 

shown by different methods of sampling. In Soil Science, v. 3, no. 6, p. 
499-514- 1917- 

2. Atwater, W. O. Cooperative experimenting as a means of studying the 

effects of fertilizers and the feeding capacities of plants. U. S. Dept. 
Agr., Special Rpt. No. 18. 1882. 

3. . Results of field experiments with various fertilizers. U. S. Dept. 

Agr., Special Bui. No. 31. 1883. 



4l6 JOURNAL OF THE AMERICAN SOCIETY OF AGRONOMY. 

4. Barber, Clarence W. Note on the influence of shape and size of plots in 

tests of varieties of grain. Maine Agr. Expt. Sta. Bui. 226, p. 76-84. 
1914. 

5. Baule, B. "fiber die Verwertung der Fehlertheorie in der Land- und Forst- 

wirtschaft. In Fiihling's Landw. Ztg., 62: 852-866. 1913. 

6. BoLLEY, H. L. Interpretation of results noted in experiments upon cereal 

cropping methods after soil sterilization. /;/ Proc. Amer. Soc. Agron., 
V. 2 (1910), p. 81-85. 1911. 

7. BucKMAN, H. O. Fertilizing the rotation. In Jour. Amer. Soc. Agron., 

5: 157-164. 1913. 

8. Bull, C. P. The row method and the centgener method of breeding small 

grains. In Proc. Amer. Soc. Agron., v. i (1907/09), p. 95-97. 1910. 

9. Carleton, M. a. Limitations in field experiments. In Proc. Soc. Prom. 

Agr. Science, 30: 55-61. 1909. 

10. Coffey, G. N. The purpose and interpretation of field experiments. In 

Jour. Amer. Soc. Agron., 5 : 222-230. 1913. 

11. , and TuTTLE, H. Foley. Pot tests with fertilizers compared with field 

trials. /;/ Jour. Amer. Soc. Agron., 7: 129-139. 1915. 

12. Coombs, G. E., and Grantham, J. Field experiments and the interpreta- 

tion of their results. In Fed. Malay States Agr. Bui., 4: 206. 1909. 

13. Cory, V. L. The use of row plantings to check field plats. In Proc. Amer. 

Soc. Agron., v. i (1907/09), p. 68-70. 1910. 

14. Dafert, F. W. Einige Bemerkungen iiber den Zweck und die Durch- 

fiihrung von Felddiingungsversuchen. In Landw. Jahrb., 32: 149-159. 
1903. 

15. Drechsler, G. Diingungsversuche auf dem Versuchsfelde des landw. In- 

stituts der Universitat Gottingen. In Jour. Landw., 28: 243-271. 1880. 

16. . Die Theorie der Dlingung und die Aufgabe der Diingungsversuche. 

In Jour. Landw., 32: 308-336. 1884. 

17. Edler, W. Versuche zur Ermittelung des Dungerbedurfnisses des Acker- 

bodens. In Jour. Landw., 46 : 349-365. 1898. 

18. Ehrenburg, p. Versuch eines Beweises fiir die Anwendbarkeit der Wahr- 

scheinlichkeitsrechnung bei Feldversuchen. In Landw. Vers. Stat., 87 : 
29^88. 1915. 

19. Farrell, F. D. Interpreting the variation of plat yields. In U. S. Dept. 

Agr., Bur. Plant Indus. Circ. 109, p. 27-32. 1913. 

20. . Basing alfalfa yields on green weights. In Jour. Amer. Soc. Agron., 

6: 42-45- 1914. 

21. Gardner, F. D., and Runk, J. A. Variety test of oats. In Ann. Rpt. Pa. 

State College, 1910/1911, p. 164. 1912. 

22. Giles, P. L. On the plans of fertilizer experiments. In Jour. Amer. Soc. 

Agron., 6: 36-41- I9i4- 

23. Grantham, A. E. The effect of rate of seeding on competition in wheat 

varieties. In Jour. Amer. Soc. Agron., 6: 124-12S. 1914. 

24. Hall, A. D. Fertilizers and manures, p. 366, 367. E. P. Dutton & Co., 

New York. 1909. 

25. . The experimental error in field trials. In Jour. Bd. Agr. (London), 

V. 16, no. 5, p. 365-370. 1909- 

26. , and Russell, E. J. Field trials and their interpretation. In Jour. 

Bd. Agr. (London), Supplement, p. 5-14. Nov., 191 1. 



ui'.roR r oi' COM M I I'l l: i-, on I'ii'.i.i) i-:.\ i'i-:i< i m pints. 



417 



J/. llANRorii. - -. Dctcnniiiatioii of piDhahlc i-nor in Ik-Id cxpcriincnts. In 
Mitt. Doutscli. Laiulw. (jcscll.. 28: 70-73, 87-(;(), 105-107. 1913. 

28. H.\RRis, J. A. On tlio si>;nil"icanco of variety tests. In Science, n. s., v. 
36, no. (JJ3, p. 318-3J0. igiJ. 

JQ. . Supplementary note on the significance of variety tests. /;/ Science, 

n. s., V. 37, no. 952. p. 493. 494. 1913. 

30. . An illustration of the influence of substratum heterogeneity upon 

experimental results. /;; Science, n. s., v. 38, no. 975, p. 345, 346. 1913. 

31. . On a criterion of sul)stratuni homogeneity (or heterogeneity) in 

field experiments. In Amer. Nat., v. 49, no. 538, p. 430-454. 1915. 

32. HoLTSMARK, G., and Larsen, B. R. Uber die Fchler, welche bei Feldver- 

suchen durch die Ungleichartigkeit des Bodens bedingt werden. In 
Landw. Vers. Stat., 65 : 1-22. 1906. 

33. Huston, H. A. Field tests with fertilizers. In Original Communications 

8th Internat. Congress Appl. Chem., 15: 139. 1912. 

34. Jardine, W. M. Methods of studying the relative yielding power of kernels 

of different sizes. In Proc. Amer. Soc. Agron., v. i (1907/09), p. 104- 
108. 1910. 

35. Kaserer, H. Beitrag zur Frage der Feststellung des Niihrstoffgehaltes 

einer Ackerparzelle. In Ztschr. Landw. Versuchsw. Osterr., 13 : 742-747. 
1910. 

36. KiESSELBACH, T, A. Studics concerning the elimination of experimental 

error in comparative crop tests. Nebr. Agr. Expt. Sta. Research Bui. 12. 
1917. 

37. LiEBSCHER, G. Untersuchungen iiber die Bestimmung Diingerbedurfnisses 

der Ackerboden und Kulturpflanzen. In Jour. Landw., 43 : 4g-So. 1895. 

38. Lyon, T. L. A test of planting plats with the same ears of corn to secure 

greater uniformity in yield. In Proc. Amer. Soc. Agron., 2 (1910), p. 
35-57-